戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for sap hana...

30
Dell - Internal Use - Confidential 戴爾企業技術 戰略架構師 架構大數據

Upload: lydat

Post on 25-May-2018

222 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

戴爾企業技術戰略架構師

架構大數據

Dell - Internal Use - Confidential

2

Dell - Internal Use - Confidential

3

Blueprints

3

3

Dell - Internal Use - Confidential

4

傳統和新 的交叉點

Dell - Internal Use - Confidential

5

我們的方法為您提供了更好的結果

一個可擴展的端到端的方法提供了

未來就緒的企業bull 經濟的端到端的解決方案bull 可以實現效率的最大化bull 可以在任何規模進行部署

bull封閉的解決方案

bull有限的互操作性

bull技術鎖定

專有系統

bull複雜的整體式系統

bull高昂的每筆交易和擴展成本

bull由單一的供應商提供專有的技術

舊式ITbull低成本的商品化組件但不具有解決方案的增值性

bull缺乏足夠的技術支持bull無法確保使用的可持續性

商品化系統

增加運營成本

降低運營成本

降低購置成本增加購置成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 2: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

2

Dell - Internal Use - Confidential

3

Blueprints

3

3

Dell - Internal Use - Confidential

4

傳統和新 的交叉點

Dell - Internal Use - Confidential

5

我們的方法為您提供了更好的結果

一個可擴展的端到端的方法提供了

未來就緒的企業bull 經濟的端到端的解決方案bull 可以實現效率的最大化bull 可以在任何規模進行部署

bull封閉的解決方案

bull有限的互操作性

bull技術鎖定

專有系統

bull複雜的整體式系統

bull高昂的每筆交易和擴展成本

bull由單一的供應商提供專有的技術

舊式ITbull低成本的商品化組件但不具有解決方案的增值性

bull缺乏足夠的技術支持bull無法確保使用的可持續性

商品化系統

增加運營成本

降低運營成本

降低購置成本增加購置成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 3: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

3

Blueprints

3

3

Dell - Internal Use - Confidential

4

傳統和新 的交叉點

Dell - Internal Use - Confidential

5

我們的方法為您提供了更好的結果

一個可擴展的端到端的方法提供了

未來就緒的企業bull 經濟的端到端的解決方案bull 可以實現效率的最大化bull 可以在任何規模進行部署

bull封閉的解決方案

bull有限的互操作性

bull技術鎖定

專有系統

bull複雜的整體式系統

bull高昂的每筆交易和擴展成本

bull由單一的供應商提供專有的技術

舊式ITbull低成本的商品化組件但不具有解決方案的增值性

bull缺乏足夠的技術支持bull無法確保使用的可持續性

商品化系統

增加運營成本

降低運營成本

降低購置成本增加購置成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 4: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

4

傳統和新 的交叉點

Dell - Internal Use - Confidential

5

我們的方法為您提供了更好的結果

一個可擴展的端到端的方法提供了

未來就緒的企業bull 經濟的端到端的解決方案bull 可以實現效率的最大化bull 可以在任何規模進行部署

bull封閉的解決方案

bull有限的互操作性

bull技術鎖定

專有系統

bull複雜的整體式系統

bull高昂的每筆交易和擴展成本

bull由單一的供應商提供專有的技術

舊式ITbull低成本的商品化組件但不具有解決方案的增值性

bull缺乏足夠的技術支持bull無法確保使用的可持續性

商品化系統

增加運營成本

降低運營成本

降低購置成本增加購置成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 5: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

5

我們的方法為您提供了更好的結果

一個可擴展的端到端的方法提供了

未來就緒的企業bull 經濟的端到端的解決方案bull 可以實現效率的最大化bull 可以在任何規模進行部署

bull封閉的解決方案

bull有限的互操作性

bull技術鎖定

專有系統

bull複雜的整體式系統

bull高昂的每筆交易和擴展成本

bull由單一的供應商提供專有的技術

舊式ITbull低成本的商品化組件但不具有解決方案的增值性

bull缺乏足夠的技術支持bull無法確保使用的可持續性

商品化系統

增加運營成本

降低運營成本

降低購置成本增加購置成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

初始成本

持續成本

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 6: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

6

Dell is different

Modular systemsNo costly monolithic stacks

Open approachNo intentionally closed ecosystems

Modern portfolioNo vested interest in legacy

systems

Flexible scalingNo forced constraints or

rip-and-replace

Standards-basedNo deliberate technology lock-in

End-to-end solutionsNo siloed viewpoint or

hidden agenda

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 7: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

7

500-12000 usermailboxes

MicrosoftSharePoint

Lync Exchange Ref Arch

End to End CCC Ref Arch from client to datacenter (up to 10k+ users)

Red Hat Openstack Ref

Arch (SM L sizing)

With ASM DCM

Ref Arch with Cloudera

Hadooop SAP Hana with Boomi

Statistica SharePlex TOAD

etc

Oracle (OLAP OLTP) amp SQL

(incl Fast Track) Ref Arch with

sizing with TOAD

SharePlex etc

HPC Ref Arch ndashNFS HSS amp Intel Edition

(IEEL) (S M L sizing)

50-1000 VMs MS Hyper-V amp VMware ESX

Ref Arch (S M L)

Cloud Platform System (CPS)

Dell Acceleration Appliance for

Databases (DaaD)

Dell IntgSolution for

Oracle Database 12c (DISOD)

XC Series

Dell Genomics Platform

Ref Arch

Engineered Solutions

XC Series

Cloudera Spark Syncsort

Analytics Platform System

You are here

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 8: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

8

Flexible solutions tailored to your organizationrsquos goals

Engineered Solutions

Dell Blueprint for Solutions

Validated and Optimized for success

Accelerated Time to Value Solutions

Simpler to deploy amp manage lower risk

Performance and Efficiency at any Scale

Exceptional Execution amp Delivery

Workstations to supercomputer clusters to the cloud

Optimize your entire ecosystem from laptop to petaflop to the cloud

with the only global end-to-end solutions provider

Reference Architectures

Best of Breed Products

Built for High Performance Computing

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 9: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

9

How data is moved and prepared

for analysis

Data integration aggregation and transformation

Where data originates

Databases

Social media

Sensor data

Devices

LOB applications

Cloud

External sources

Where data is analyzed

Analytical engine

Business intelligence

In-memory computing

Enterprise data warehouse

大數據和分析的基礎知識

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 10: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

10

Hadoop 架構

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 11: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

11

Hadoop 架構

bull The 2U high rack-based FX2 converged infrastructure enclosure can host different blocks of compute and storage resources ndash depending on workload needs

bull Converged infrastructure provides efficiency of shared power networking IO and management as well as greater overall density

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 12: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

IT和業務保持一致bull 界定目標

bull 設置關鍵績效指標(KPI)

bull 評估環境

bull 预測需求

業務變成由數據驅動

革新業務bull 擁抱智能bull 打造分析能力bull 交付敏捷性和安全的商務

智能

提高運營效率bull 適應不斷變化的需求bull 優化績效bull 降低成本

大數據之旅牢記以終為始

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 13: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

13

大數據成功的關鍵

願景 70

文化

組織

流程

業務 20

應用集成

數據管理

技術

Hadoop 10

數據模型

分析

BI

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 14: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

14

戴爾的大數據方法論

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 15: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

15

戴爾大數據及物聯網架構

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 16: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

16

如何開始大數據的戰略第一步

戴爾中國大數據聯盟創新實驗室httpbigdatademocn

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 17: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

17

宏觀概述 客戶使用場景

Dell Syncsort Cloudera

RA

Tim

e f

or

An

aly

sis

Seconds

Operational Efficiency(OE) Use Cases

Minutes

10s

Seconds

10s

Minutes

Hours

DWFT

SAP HANA

Microsoft APS

Business Transformation(BT)

Use Cases

Cloudera In-Memory Appliance

Cloudera RA

Dell StatisticaAnalytics

De

ll So

ftwa

re (D

S)

Po

rtfolio

Dell Toad Data PointIntelligence Central

Dell Boomi

Solutions Can Have Multiple Blue Print Components Per Use Case

Engineered Solution

Reference Architecture

Dell Software

Data Management

Data Integration

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 18: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

18

Old Way New Way with Hadoop

bull Built Around RMDBEDW

bull High SW Costs

bull Structured Data Only

bull More Transactions in DB = Slower Performance

The Result

bull Augment The Database

bull Lower SW Costs

bull All Data Types

bull Move costly workloads into Hadoop

bull Drive Operational Efficiency

bull Lower Cost To Store Data

bull Lower Data Transformation Cost

DB

Data Sources

Data Staging

Clean amp Parse Data

ETL

BI

Query Reports Data Native Format

Data Sources

ETL

Da

ta

Disc

ov

ery

An

aly

tics

Data Driven Business

Hadoop amp Analytics

ETL卸載是在大數據之旅的第1步補充了傳統的工具並引進Hadoop作為一個技術以降低運營成本

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 19: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

19

太多工作負載在

Traditional Data Pipeline

Enterprise data warehouse + ETLData Transformation Jobs

Business ReportingQuery

Data Staging ToolExtract amp Load DataClean amp Parse Data

Disparate Data Sources

The Results Longer data transformation job

times

Not meeting SLAs for business reporting

Slow Ad Hoc Query

Too costly to scale

Perf

Capacity

Modern Data Pipeline

Disparate Data Sources

Hadoop + ETLData Transformation JobsClean Parse Transform

Enterprise data warehouseBusiness Reporting

Query

Capacity

Perf The Results Reduced data transformation

job times

Improved SLAs for business reporting

Fast Ad Hoc Query

Scales Economically

用Hadoop使數據管道現代化

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 20: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

20

Hadoop Connectivity

Hadoop ETL

Hadoop Sort

bull Smarter Architecture ndash ETL engine runs

natively within MapReduce

bull Smarter Connectivity ndash One tool to

connect all your data even mainframe

bull Smarter Development ndash Hadoop ETL

without coding

bull Smarter Productivity ndash Use Case

Accelerators for common ETL tasks

bull Smarter Security ndash Enterprise-grade

security

PLUS Smarter Hadoop

bull Enhanced vertical scalability

bull Smart contributions to open source

community

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 21: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

21

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 22: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

22

SILQ is a key differentiatorbull SILQ provides integrated analysis of SQL

jobs and queries and generates a graphical data flow for DMX-h

bull This web based utility helps you shift ELT processing from the data warehouse into Hadoop

bull A ldquono codingrdquo approach complex SQL is replaced with a powerful easy-to-use graphical data flow

bull Provides best-practices to develop DMX-h jobs

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 23: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

23

With SILQ

5 hours

Without SILQ

20 hours

Example Analyzing 1 Script with 1000 lines of Complex SQL Code

Let Your EDW Offload Project Take Flight

Upload SQL file

Hit lsquoChartrsquo amp Visualize

Click functional blocks

Review recommendations

Understand SQL code

Outline steps + work flow

Categorize specific tasksjobs

Manually update code

Manual Syncsort

10 Jobs $15000 $375

100 Jobs $150000 $3750

500 Jobs $750000 $18750

97 Savings

Estimates at $75 hour

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 24: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

24

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 25: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

25

Syncsort DMX-h deploys and runs as part of your Hadoop cluster

bull Automatically adapts and optimizes for hardware sources

bull Delivers faster throughput per node

bull No code to develop maintain or tune

Full integration with Cloudera Manager

bull Deploy monitor amp manage large-scale Hadoop clusters

Syncsort contributions to Apache Hadoop ship on CDH

bull Pluggable Sort Sqoop Integration and more

Best-in-class mainframe data ingest capabilities

bull Secure mainframe data access through SFTP amp connect direct

bull Spark connector for mainframe

Integration with HCatalog

bull Facilitates metadata management amp data lineage

No-hassle support for Kerberos amp LDAP

bull Secure your new environment including authenticated sampling amp browsing

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 26: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

Microsoft Analytics Platform System by Dell

bull Integrated compute storage networking and software appliance for high performance database workload needs

bull Microsoft APS software aggregates stores and queries relational (SQL)+ non-relational (Hadoop) data in the solution

bull Includes Jumpstart services (3 weeks) for customer training and architecture design

The Dell Difference

bull MPP (Massively Parallel Processing) appliance for up to 100x improvement over SMP database workloads

bull Highly scalable solution ndash starting from 3 nodes to 54 Multiple racks can be configured (up to 6 racks) Scales from 21TB to 6PB Scale-out expansion 3 nodes at a time

bull White glove delivery and installation Delivered as fully built appliance with software installed and configured for the customer with training services

Link to APS Solution BriefLink to Tech Sheet

Blueprintfor Big Data amp Analytics

Engineered Solution Microsoft Analytics Platform System by DellReal-time management of relational (SQL) and non-relational (Hadoop) data

x2 | SX6036 Infiniband Switchesx2 | N3048 Ethernet Switchesx2 | R630 Management Nodesx2 | R630 Nodes added when HDInsight included in first rack Optional

3rd Scale Unite for 9 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

2nd Scale Unite for 6 Nodes Optionalx3 | R630 Compute Nodes x2 | MD3060e JBODs (102 Drives 18 Spare)

Base Unit for 3 Nodesx3 | R630 Compute Nodesx2 | MD3060e JBODs (102 Drives 18 Spare)

Scales from 3 nodes to 54 nodes across 6 racks (up to 6PB)

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 27: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

Blueprintfor Big Data amp Analytics

Massively Parallel Processing Real-Time Decision Support of Relational and Unstructured Data

1 SOURCE

Sensors

Customer Data

Order Data

Asset Tracking

ERP

2 INTEGRATE AGGREGATE amp TRANSFORM

MANAGEMENT

Azure IoT FabricEvent Hubs

SECURITY DESIGNDEPLOY

3 ANALYZE

4 ACT

Predictive Analytics Machine Learning

Polybase(nMicrosoft ative

in APS)

In-memory relational ampnon-relational

harmony

Microsoft APS by Dell

Relational data aggregation

On-Prem amp Cloud Options

Unstructured data aggregation

High Speed ETL

High Speed ETL

SERVICES

In Memory

StatisticaAnalytics

A

B

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 28: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

28

Scale-out configurations

bull Multi-node

bull Scalable and highly available

bull SUSE or RedHat Linux for SAP support

bull Easy non-disruptive expandability

bull Compellent fibre channel SAN for enterprise class features

bull Built to be mission critical

Single server configurations

bull Self contained pre-configured sizes

bull SUSE or RedHat Linux for SAP support

bull Optional vMWare vSphere virtualization

bull Up-to 3TB RAM for Biz Suite on HANA

Grow from 2TB to 24TB -without disruption ndash

in 1 or 15TB increments

ScalableHighly Available

2TB HA 24TB HA

2TBBusiness Suite

4x Intel E7v3

128GB 256GB 512GB

2x Intel E7v3

1 15TB

4x Intel E7v3

TDI Intel 2S configurations

bull 2S Tower rack blade options

bull Ideal entry point for devtestpoc

bull Certified for TDI deployment

R730 R730xd

Intel E5-v3Rack mount2U form factor

R630 M630

Intel E5-v31U Rack orfrac12 slot blade

T630

Intel E5-v3Tower designDrive density

3TBBusiness Suite

4x Intel E7v3

Ready for SPS09 multi-tenancy and dynamic tiering

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 29: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

29

Building a complete managed infrastructure

Dell Engineered Solutions

for SAP HANA

Application

Database

OS

Virtualization

Servers

Network

Storage

Service Management

Disa

ster R

ec

ove

ry Pro

ce

ss

Backup amp Recovery

Mo

nit

ori

ng

amp M

ain

ten

an

ce

(Pre

dic

tive

Op

era

tio

ns

Man

ag

er)

FoglightService management storage performance resource

planning amp optimization change managementActive System Manager

Provisioning capacity on demand consolidation

(coming soon)

FoglightPerformance management amp

monitoring network monitoring

Toad for SAPDatabase management development tuning amp

analysis

SAP Solution ManagerSystem monitoring business process monitoring central system administration SAP

early watch alert

Shareplex for SAP migrationproduction data replication

SAP System replicationserver based SW replication

DR 6000 NetBackup AppliancePurpose-built backup-to-disk appliance scalable wsource-

side ingest protocol accelerators

System ReplicationSynchronous asynchronous

storage replication

Duplicate appliance resources at DR site and use Compellentreplication services to synch storage and SAP replication services to synch processes

DR system can also be used for testdev by deploying additional

DASD or SAN disk resources Data synch continues on main DR SAN but server operate as testdev using secondary disk

resource ndash until failover is required

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward

Page 30: 戴爾企業技術 戰略架構師 - veritas · resources –depending on workload ... for SAP HANA Application Database OS ... Toad for SAP Database management, development, tuning,

Dell - Internal Use - Confidential

30

Let Dell help youchoose the right path forward