Database Disaster Recovery

Document Sample
Database Disaster Recovery Powered By Docstoc
					Database Disaster
Recovery
Tom Hamilton, NetApp
America’s Channel Database
CSE




                             1
Agenda

 Definitions and Issues
 Oracle Data Guard
 Microsoft SQL Server Mirroring
 NetApp SnapMirror
 MetroCluster




                                   2
        Customer Challenges

                               Power Failure                                                       42%
Internal
Data Center              IT Hardware Failure                                            31%

Failures                     Network Failure                                      21%

                          IT Software Failure                               16%

                                Human Error                                 16%

                                       Flood                          12%

                                   Hurricane                        10%

External                                 Fire                  7%
Data Center
                               Winter Storm                   6%
Failures
                                   Terrorism             4%

                                 Earthquake            3%

                                    Tornado         2%

                               Chemical Spill     1%



   Source: Forester / Disaster Recovery Journal Global Disaster Recovery Preparedness Online Survey, Oct, 2007
                                      NetApp Confidential - Limited Use                                    33
    What is Data Guard?
 Data Availability & Data Protection solution for Oracle
 Automates the creation and maintenance of one or more
  synchronized copies of the primary database
 If the primary database becomes unavailable, a standby
  database can easily assume the primary role
 Standby databases can be used for queries, reporting,
  testing, or backups while in standby role
 Feature of Oracle Database Enterprise Edition (EE)




                                                            4
      Oracle Data Guard Architecture

                                              Dallas            Physical    Open R/O
                                                                Standby
                        Sync or Async
                        Redo Shipping
 Production                                             Redo
 Database                                               Apply
              Network
                                                                            Backup
                                    Broker
                                                                             DIGITAL DATA STORAGE




                                                                                  DIGITAL DATA STORAGE




                                                        Logical             Open R/W
                                        Transform       Standby
Chicago                                 Redo to SQL




                                                SQL
                                                Apply
                                                                           Boston


                                                                                                         5
  Flexible Data Protection Modes

   Protection Mode          Risk of Data Loss         Redo Shipment

  Maximum Protection          Zero Data Loss          Synchronous redo
                          Double Failure Protection       shipping

  Maximum Availability        Zero Data Loss          Synchronous redo
                          Single Failure Protection       shipping

 Maximum Performance            Potential for         Asynchronous redo
                             Minimal Data Loss            shipping



Balance cost, availability, performance, and transaction protection




                                                                          10
   Switchover and Failover

 Primary and Standby role transitions
 Switchover
  – Planned role reversal
  – No database reinstantiation required
  – Used for maintenance of OS or hardware
 Failover
  – Unplanned failure (e.g. disasters) of primary
  – Use Flashback Database to reinstante original primary
 Manually execute via simple SQL / GUI interface, or
 Automate failover using Data Guard Fast-Start Failover


                                                        11
 Active Data Guard 11g
 Increase ROI - Real-time Query


                                                 Real-time
                                                  Real-time
                                                 Query
                                                  Queries



                         Continuous Redo
                         Shipment and Apply



           Production                     Physical Standby
           Database                       Database

• Offload read-only queries to physical standby
• Offload fast incremental backups to physical standby


                                                              14
Microsoft High Availability

 Failover clustering
 Database mirroring
 Log shipping
 Replication




                              15
Microsoft SQL Server Database Mirroring

 Benefits
 – Increased data protection
 – Increased database availability
 – Database availability during upgrades
 How SQL Server Mirroring Works
 – Operating modes
     High safety mode
     High performance mode




                                           16
SQL Server Database Mirroring




        High safety mode


                                17
SQL Server Database Mirroring




        High performance mode



                                18
                  SnapMirror – Our Flagship DR Technology

                                           Value Proposition
Primary Data Center
                                           Simple, flexible and cost-effective

 LAN                                       Benefits
                                          ■ Simple
                                             –Simple configuration
                                             –Integrated with SnapManager
                FAS                FAS       –Simplified failover/failback - MultiStore
                                          ■ Flexible
                      SnapMirror             –Address a broad range of DR requirements
DR Site                                      –Operate with FC or IP network
                                             –Mirror to/from any NetApp system
                                             –Multi-hop, cascading
                                          ■ Cost-effective
                                             –Mirror to inexpensive targets
          FAS                                –Supports all SLAs
                                             –Bandwidth efficient with BLI changes
                              NearStore
                                             –Leverage low cost IP networks
                                             –Backup data can be made writeable
                                                                                          19
     SnapMirror Flexibility
Synchronous SnapMirror
            1       Every Write             2        Zero data loss
                                                     Distance limited
                                                     Performance impact
            4               3


Semi-Synchronous SnapMirror
                                                     Small data loss
         1           Every Write            3
                                                     No distance limit
                                                     No performance impact
            2              4


Asynchronous SnapMirror
                                                     Data loss > 1 minute
        1                                   3
                                                     No distance limit
                A      1
                           Changed blocks
                                                B    No performance impact
                       2
                            Set intervals
        2
                                                                              20
     Flexibility

 Multiple hops                                                               Many-to-one, One-to-many

                         Cascading



                 Async               Async

     Sunnyvale           New York            London



 Asymmetric replication
                                                      FAS6080                              FAS3040




 Heterogeneous
  replication with V-Series
                                                                       V-Series
                                                       Enterprise                        FAS
                                                       Storage Array

                                                                                                         21
      NetApp Solution:
      Rapid Recovery for Business Apps
                                                            Replication
 Primary Data Center                 DR or 2nd Datacenter
                                                            ■ SnapManager coordinates
                                                              consistent snapshots

                                                            ■ SnapMirror replicates
                                                              snapshot copies – ensure
  Oracle Server                                               restartable copies

                       SnapManager
                                                 Failover   ■ Mirror database and logs
                                                 Server       at different frequencies
                       SnapDrive™




                         Database                FAS
FAS
System                                           System
                          Logs

                       SnapMirror®



                                                                                    22
Database Mirroring Vs. SnapMirror

Data transfers
Licenses
Server-server vs. controller-controller
Reversible sync
Failover
Database only vs. everything
Go for win-win




                                          23
     MetroCluster Delivers Continuous
     Availability
MetroCluster combines array-based clustering with synchronous mirroring to
provide zero data loss with zero downtime

                                            Set-it-once simplicity

                                            Zero change management

                                            Zero planned downtime

                                            Zero unplanned downtime


                                            Up to 80% read performance
                                             improvement

                                            50% lower cost and complexity
                                             than our competitors

                                                                             24

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:2
posted:2/25/2013
language:Latin
pages:19