Learning Center
Plans & pricing Sign in
Sign Out

Building Highly Available Infrastructure _ Non-disruptive Backup


									Building Highly Available Infrastructure & Non-disruptive Backup & Recovery for SQL 2005
김영택 차장( EMC Korea

Today’s SQL Server Challenges
How do I consolidate instances? How do I migrate onto another storage platform? How do I accelerate implementations?

How do I minimize downtime?

What service levels can I promise my customers?

How do I minimize complexity? How do I optimize performance?

Where do I find the resources?

EMC’s Information Lifecycle Management
A Customer Strategy to Align IT Infrastructure with the Business, Based on the Changing Value of Information over Time

Reduce Total Cost

Improve Business Value



Intelligence Move Secure

Key Customer Initiatives: EMC ILM Solutions
Phase 1: Tiered Infrastructure
App App App

Phase 2: Application-Specific ILM
App App App

Phase 3: Cross-Application ILM
App App App






Classification and Policy Consolidation Continuity Reduce Total Cost Backup-to-Disk

Comprehensive Enterprise Content Backup, Recovery, Management and Archive Policy-based Compliance Automation Content Management Improve Business Value

EMC - The Trusted, Open, Information
Infrastructure Provider
 $2.243B in revenue in Q1 2005
 $1,026M in systems (~46%)  $832M in software (~37%)

“EMC’s diverse

portfolio of storage  $375M in Storageservices (~17%) Storage &assets spanning Information hardware, software,  Companyin a row that EMC has 10th quarter Information Infrastructure met or exceeded targets Mgmt and services leaves it in the best overall  Q1 revenue growth of 20% position in the industry.”  Double-digit revenue growth in

1991-2002 every geography


— Goldman Sachs


 ~$1B R&D in 2006, over 2,000  Nearly 24,000 employees

EMC’s Partnership with Microsoft
 Partners since 1995  Working together to create the

leading products, services, and solutions
 Project Real - Joint effort to

Mutual Product Usage

Complementary Engineering

discover best practices for BI applications on SQL Server 2005
 Schema design including partitioning  Management and maintenance of systems  Management of incremental data updates

Cooperative Support and Services

Joint Solution and Best Practices

 Customer benefits:
 Leading storage solutions on the Microsoft Environment — now and in the future  Increased ROI

Project REAL

EMC Solution Suite for SQL Server
Customer Challenges
 Upgrade to SQL Server 2005  Minimize disruption  Lower total cost of ownership  Simplify infrastructure  Downtime  Meet backup windows

EMC Solutions
Upgrade and Migration S Q L S E R V E R

Comprehensive Backup and Recovery

 Long recovery times  Eliminate impact on application  No data exposure
 Efficiently repurpose data  Data-access impact on production  Limited resources

Disaster Recovery Data Replication for Operational Productivity

Consolidation with SQL Server
Instance Consolidation Value After Before
App SQL App SQL App SQL App App App App App

 Higher Asset Utilization
 Servers  Storage



 Lower Operating Costs
 Tiered storage  Less to manage  Lower environmentals

Windows O/S

 Higher Availability
Instance Consolidation SAN Consolidation
 RAID arrays  Channel failover

Dedicated Servers & Storage

 Enables Higher SLAs
 Server clustering  Rapid recovery options

EMC Tiered Storage Platforms
The Choice, Flexibility, and Integration Required to Build a Solid Foundation
Integrated Management Advanced Data Replication

Wide range of storage platforms, networks, replication, and management options

Block Access

File Access

Object Access

Tiered storage based on service levels and business requirements

High End
CX500 CX700 NS Series / Celerra Integrated CNS


Entry Tier

DMX1000-M2 DMX2000-M2 DMX3000-M2

Fibre Channel Disks
CX300 AX100

Serial ATA Disks
Centera DMX800 DMX1000 DMX2000

DMX3000 DL300 ADIC Scalar Family

Designed to work together to minimize operational effort

NetWin NS Series / 110 / 200 Gateway





Symmetrix DMX
High-end Storage Platforms Today
 Proven Direct Matrix

 Increased scalability
– Cost savings through

– Highest service levels – Predictable performance

massive consolidation

 Mission-critical

 Increased modularity
– Grow performance and

– Non-disruptive everything – Designed for continuous operations

capacity when needed

 Tiering ―in-the-box‖
– Multiple service levels

 Scalable performance
– Optimize performance for application requirements – Scales with growth

without multiple arrays

 Self-optimizing
– Less administrative effort

 Advanced replication
– Protects critical data without impacting application performance

 Investment protection
– Ease migrations to new


Mid-tier Storage Platforms Today
 Proven CLARiiON

 Simplified

– Cost-effective performance

– Wizard-based tools

and availability

 Modular design
– Protects investment

 Advanced software
– Protects information
– Simplifies management

guide users through common tasks – Easier to deploy and implement functionality

 More functionality  Application integration  Performance and

 Ease-of-use  Flexible choices
– Match config to requirements

– Minimizes training AX100 CX300s X500s CX700s CX500i CX300i AX100i


NAS: EMC Celerra
Network Attached Storage Today
 Enterprise class NAS
– Highly scalable – Highly available – Multi-protocol

NS700 Celerra CNS

 Increased scalability
– Larger file sizes, larger file systems – Scale, capacity, connectivity, or performance

 Flexible configurations
– Gateways for Sym and CLARiiON – Integrated NAS appliances

 Global accessibility
– Centrally manage distributed data
NetWin 110 NetWin 200 NS700G

 Flexible management
– Single console manages multiple servers

 Policy-based data

 Unified name space
– Simplify access, management – Abstraction of the physical backend environment

– Automates data placement – Reduces operational costs – Transparent to client applications

Protection & Recovery Options
Local Remote
Disk based Point-in-Time (PiT) Copy Synchronous Remote Mirroring. Asynchronous Remote Mirroring PiT Copy

Service Levels

BCV/ Clone PiT Copy Snapshot Backup to Disk


Recovery Continuum

Local Protection Solutions for SQL Server
Sample Architecture
IP network
Database Servers Fibre Channel

Solution Features

Backup-to-clone option
Backup Server (Networker)

 Online backup  Restore in minutes  EMC Replication Manager
Integration with VSS for SQL 2005 Integration with VDI for SQL 2000 and up

Snap or Clone Backup to clone

Tape images
or Disk images

Backup-to-disk option
 Online backup  Restore in hours  Low cost ATA disk  Option for SAN, NAS  Disk or tape image

Backup-toclone option

Backup-todisk option

Accelerate deployment
 EMC Consulting

Simplified Instant Local Recovery
Without EMC:
 Tape backup may be tolerable, but tape restore is not  Volume Shadow Copy Services (VSS) is widely touted, but few solutions are completely MSsupported  VSS Solutions are complex to manage

With EMC:
 Simple instant VSS backup and restore  Simple database & table recovery  Augments and integrates with existing tapebased backup  Most flexible replication capabilities  MS supported VSS solution

Recovery source EMC Replication Manager and SnapView Tape

Recovery time

Notes Restore time Log playback time (up to 80% faster)
Multiple hours


Symmetrix Remote Protection Solutions for SQL Server
Servers Source Target

• No data exposure • Some performance impact • DMX: SRDF-S or CE
Servers Source

Limited Distance


• Seconds / hours of data
• •
exposure No performance impact DMX: SRDF-A
Servers Source

Unlimited Distance

Asynchronous Point-in-Time

• Hours / day of data exposure • No performance impact • Requires BCVs, clones, or snaps • DMX: TSIM / RM / SRDF-DM

Target Unlimited Distance


CLARiiON Remote Protection Solutions for SQL Server
Servers Source Target

• No data exposure • Some performance impact • CX: MirrorView/S
Servers Source

Limited Distance


• Seconds / hours of data
• •
exposure No performance impact CX: MirrorView/A
Servers Source

Unlimited Distance

Asynchronous Point-in-Time

• Hours / day of data exposure • No performance impact • Requires BCVs, clones, or snaps • CX: RM / RMSE / SANcopy

Target Unlimited Distance


Faster Recovery from Disaster
Replication method Recovery-Point Objective (RPO)
Potential data loss

Recovery-Time Objective (RTO)
Time to recover full service


No loss Minutes

Seconds to Minutes Minutes to hours

Asynchronous point in time

Hours / Day

Minutes to hours

Options for cost versus target RPO and RTO

Replication Manager
 Software that dramatically simplifies and

automates the creation, management, and usage of disk-based replications  Integrating applications, hosts and arrays  Scaling from entry-level to enterprise and enabling both storage-savvy personnel and storage novices  Automates the discovery process
 Databases and filesystems  Available storage

Foundation of Replication Management Usability • Point-andclick • Flexibility • Recovery Intelligence • Application • Array • Host

 Understands database layout works

APIs to create recoverable images  Maps host to storage to replicas  Wizard driven process creation

Heterogeneous • Applications with • Arrays • Hosts

Automation • Discovery • Process execution • Scheduling • Cataloging

Fast & Reliable Information Access




Replica 1



Replica 2



Replica 3



Disk-based replicas save time


FullTime AutoStart
Increasing Application Uptime
Reality: System maintenance or component outages are common, both causing downtime…

AutoStart manages automated application and data restart on an alternate server, either local or remote.
SRDF/A, SRDF/S or RepliStor


AutoStart seamlessly transfers control of application and storage resources to a local or remote facility in the event of an outage. AutoStart provides functionality to assist in automating fail back of services, apps and data.


Target Environments
 


Oracle 9i/10g on Solaris, HP-UX, AIX, Windows Exchange, SQL, Print Service on Windows


EMC Replication Support
  

SRDF/A, SRDF/S MirrorView/A. MirrorView/S RepliStor

Protects Against Local and Remote Failures
Local cluster with mirrored storage
 Zero data loss – data mirrored over IP  No external array required  No shared storage required  Low-cost hardware-availability solution
Primary server Network Dedicated link Mirrored Standby machine

Local cluster with shared storage
 Data shared via SCSI or SAN  Ideal hardware-availability solution for

data center or campus environments
Site A

Geographically dispersed cluster


Site B

with remote data replication
 Long-distance data replication
 Protect against site disaster
SRDF/S, SRDF/A MirrorView/S, MirrorView/A, RepliStor

Comprehensive Backup, Recovery, and Archive
Identify and Archive Infrequently Used Data
For e-mail, databases, file systems, content repositories EmailXtender DatabaseXtender  Smaller production environments DiskXtender  Less primary storage Documentum CSS  Less server resources  Easier management

Reduced Backup Volumes
Enables use of disk as backup medium • • • • Faster backup Faster recovery Less cost Less effort

Active Archive
Disk Keeps data protected, accessible, and compliant • Simplified management • Easier access • More value from information


CLARiiON Disk Library Centera CLARiiON / ATA CLARiiON ATA Celerra / ATA Plan • Build • Manage • Support Celerra ATA NetWorker Experienced professionals help define service levels ADIC Tape Library and speed deployment • Consistent SLAs • Measurable processes Information Protection Services • Faster deployment Data Classification Services

Archiving Solution for SQL Server 2005
 Partition by date (example), store by LUN
Q2 2004 Q3 2004 Q4 2004 Q1 2005 Q2 2005 Q3 2005 Q4 2005

 Move to lower cost storage

via EMC LUN migration  Policy-based, Automated
Q2 2003 Q3 2003 Q4 2003 Q1 2004 Q2 2004

 EMC Custom Services

Tiered Storage with SQL Partitioning

Last quarter data is aged to higher density storage Older data is aged to highest density storage after one year

LUN Migration reduces ETL process complexity and increases storage and database availability

EMC Services Suite for SQL Server
EMC Solutions
Upgrade and Migration S Q L

EMC Services
EMC Upgrade and Migration for SQL Server EMC Database Layout for SQL Server



Comprehensive Backup and Recovery

EMC Recovery Optimization for SQL Server (local)
EMC Recovery Optimization for SQL Server (remote) EMC Replication Accelerator for SQL Server

Disaster Recovery Data Replication for Operational Productivity

EMC SQL Services Implementation
Assessment / Planning Design / Recommendations Implementation / Testing / Validation

Phase 1: SQL Data Layout Accelerator
Review current infrastructure—server, operating systems, database Review current database-I/O usage Define requirements for new MS SQL

Phase 2: SQL Data Layout Accelerator and SQL Recovery Optimization Service
Provide recommendations for optimal deployment of SQL Server on EMC SAN infrastructure, based on EMC best practices Provide recommendations for Local and Remote BC/DR architecture for SQL environment

Phase 3: Joint Testing and Validation (optional)
Assist your staff with the implementation, testing, and validation of recommendations Provide knowledge transfer to your staff

environment Determine recommended architecture to optimize performance, scalability, and availability

Upgrade and Migration for SQL Server
 Provides an accelerated a SQL Server Migration and/or

Upgrade onto an EMC SAN platform  Integrated approach addressing both the SQL Server database upgrade and the infrastructure refresh planning  This solution is comprised of the following:
 Assessment, planning, design, implementation for Upgrade and

Migration  Upgrade, Database Layout, Automated Replication
 Custom upgrade plan
TCO Without EMC

 Deliverables
 Custom implementation
SQL Lifecycle Stages

With EMC

Plan and Analyze

Deploy Production

Maintain and Improve

EMC Recovery Optimization Service for SQL Server
Best practices led, business continuance design service for local and remote replication
SQL Server/Cluster Backup and Server Recovery Failover Methodology
• Provides a methodology recommendation for backup and recovery of: SQL Server/Cluster, Single Database, Single Instance, select instances, single database, multiple databases, multiple instances, deleted record, deleted table (where achievable) • Recommendation for a local replication architecture which includes software (ex. Replication Manager) and hardware required, for establishing local replication at the customer site. • Recommendation for a remote replication architecture which includes software (ex. SRDF or SAN Copy) required and hardware required for establishing a remote replication environment for disaster recovery purposes SQL Recovery Methodology Document Local Replication Architecture Recommendation Remote Replication Architecture Final Report Final Presentation outlining BC Architecture

Customer Deliverables

Databas Local Replication e Architecture for SQL Failover Server Databases

Data Center Failover

Remote Replication Architecture for SQL Server Databases

EMC Value for SQL Server Environments
Applications’ Lifecycle Demands

EMC Delivers…
   

Accelerate Optimize Manage

Faster implementations Better performance Lower risk Business continuity and disaster recovery  Manageability  Scalability  Alignment of your IT infrastructure with the changing value of information - at the lowest cost

To top