1
Oracle Data Guard 12c Zero Data Loss at Any Distance Joseph Meeks Director of Product Management, Oracle Madhu Tumma Technology Director, J P Morgan Chase 2
Program Agenda Zero Data Loss Disaster Protection - The Basics Challenge / Compromise Solution JPMorgan Chase & Co - Walk the Walk 3
Basic Principle of Zero Data Loss Requires Synchronous Communication Commit Ack Network Ack Commit Network Send Primary Standby The greater the distance, the larger the performance impact 4
Geographic Separation and Disaster Recovery DR Deployed at Distances Too Far for Synchronous Replication Typical Distance Between Primary and DR sites Greater than 300 miles Less than 100 miles 100-300 miles 233 Responses in an IOUG Survey - 2012 5
Forces Compromise Pick your poison: Sacrifice performance for zero data loss Deploy storage forwarding solutions Accept data loss and more downtime 6
Trade-off Performance for Zero Data Loss Customer Performance Tests RTT Latency Network Round Trip milliseconds Data Guard Primary Database Performance Impact Remote Mirroring Primary Database Performance Impact 0 4% 3% 10 4% 26% 15-20 10% 39% http://www.oracle.com/technetwork/database/features/availability/dataguardremotemirroring-086151.html 7
Deploy Storage Forwarding Solutions EMC Extended Distance Replication: An Example SRDF/S SRDF/A Synchronous storage mirroring to intermediate Symmetrix array Asynchronous forwarding to remote Symmetrix array for disaster recovery http://www.emc.com/collateral/hardware/technical-documentation/300-006-714-srdf-zero-data-loss-solutions-ext-distance-replication.pdf 8
Storage Remote-Mirroring: Fundamental Flaw Zero Oracle Awareness, Mirrors Every Write for Real-Time Protection Primary Database Oracle Instance (in memory) Remote Volumes Recovery files Remote mirror all changed blocks SYNC or ASYNC Recovery files Oracle data files Oracle data files 9
Deploy Data Guard Multi-Standby A Better Solution Primary SYNC Local Standby ASYNC Remote Standby Local synchronous standby HA failover with zero data loss Remote asynchronous standby for geo protection Data Guard Real-Time Cascade - new with Oracle Database 12c Many advantages: Best data protection, multiple recovery options, planned maintenance, testing 10
Data Guard: Fundamentally Best for Oracle Data Oracle Aware - Simple, Efficient, Physical Replication Primary Database Oracle Instance (in memory) database redo SYNC or ASYNC Standby Database Oracle Instance (in memory) Recovery files Recovery files Oracle data files 95% less network I/O 85% less network volume Knowledge of Oracle redo and block structures used to prevent corruption Oracle data files 11
Multi-Standby Configuration Not Ideal for All Requirements If zero data loss failover to the remote DR site is the only requirement, then multi-standby configuration would: Increase hardware, software, and support costs Increase downtime for two-step failover to the remote DR site Step 1: Failover to local standby Step 2: Switchover remote standby where production will run 12
Result: Accept Data Loss and Downtime The Compromise Chosen by Many Enterprises for Long-Distance DR Asynchronous transport Data Guard Maximum Performance Least cost with fast, single step role transition Failover means data loss Data loss poses a dilemma that increases downtime: What data will I lose if I failover? How will lost data impact the business? Instead of failing over, can I repair the problem and avoid data loss? 13
Data Guard 12c Far Sync Putting an End to Compromise 14
Data Guard Far Sync Transformation: From Compromise to Zero Data Loss Protection Primary ASYNC Remote Standby Start with any existing Data Guard configuration In this example Data Guard Maximum Performance/ASYNC with remote standby 15
Data Guard Far Sync Transformation: From Compromise to Zero Data Loss Protection Far Sync Primary ASYNC Remote Standby Deploy a light-weight Far Sync instance on a separate server Oracle instance, standby control file, redo and archived redo log files No user data files, no media recovery Locate the Far Sync instance at the maximum distance from the primary where performance SLAs can tolerate the impact of synchronous redo transport 16
Data Guard Far Sync Transformation: From Compromise to Zero Data Loss Protection Primary SYNC Far Sync ASYNC Remote Standby Configure the primary to ship redo synchronously to the Far Sync instance Configure Far Sync to forward redo asynchronously to the remote standby Supports up to 29 remote destinations Can also compress redo to conserve WAN bandwidth (requires ACO license) Standard RMAN deletion policies automate archive log management An alternate Far Sync instance on a separate server can be used for HA (not shown) 17
Data Guard Far Sync Primary Failure No Problem Primary Far Sync ASYNC Remote Standby Far Sync completes sending redo for all committed transactions 18
Data Guard Far Sync Zero Data Loss Failover Primary Far Sync Single-Step Role Transition New Primary Far Sync confirms that the configuration is in zero data loss state Fast, single-step, zero data loss failover to remote standby Same failover command as for any Data Guard configuration Manual or automatic using Data Guard Fast-Start Failover Same transparency during planned switchovers for testing or maintenance 19
Far Sync Pre-reqs Compatibility with other platforms in the Data Guard configuration My Oracle Support Note 413484.1 e.g. Far Sync for an Exadata configuration can be deployed on any Linux platform, or even on Windows Minimal storage requirements You decide how much storage to allocate for archive logs should transmission be interrupted resolve gaps without impacting primary Minimal CPU, memory, or I/O No recovery, no data files, production never runs on the Far Sync instance 20
Far Sync Instance Creation Simple and Lightweight Create a control file using a mounted primary SQL> ALTER DATABASE CREATE FAR SYNC INSTANCE CONTROLFILE AS '/tmp/fs1.ctl'; Copy the server parameter file (SPFILE) and password file used by the primary database Several initialization parameters are also modified for Far Sync Copy the control file to the Far Sync instance Create standby redo log files at the Far Sync instance Start the Far Sync instance DATABASE_ROLE column in V$DATABASE will show FAR SYNC 21
Redo Transport Details Without Far Sync Maximum Performance, all Failovers Result in Data Loss PROD LOG_ARCHIVE_CONFIG='DG_CONFIG=(PROD, DR)' LOG_ARCHIVE_DEST_2='SERVICE=DR DB_UNIQUE_NAME=DR VALID_FOR=(PRIMARY_ROLE,ONLINE_LOGFILE)' LOG_ARCHIVE_DEST_STATE_2=ENABLE DR LOG_ARCHIVE_CONFIG='DG_CONFIG=(PROD, DR)' LOG_ARCHIVE_DEST_2='SERVICE=PROD DB_UNIQUE_NAME=PROD VALID_FOR=(PRIMARY_ROLE,ONLINE_LOGFILE)' LOG_ARCHIVE_DEST_STATE_2=ENABLE 22
Add Far Sync and Designate DR as Alternate Maximum Availability Zero Data Loss FS1 Ship to DR DR LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS1, PROD, DR)' LOG_ARCHIVE_DEST_2='SERVICE=DR DB_UNIQUE_NAME=DR VALID_FOR=(STANDBY_ROLE,STANDBY_LOGFILE)' LOG_ARCHIVE_DEST_STATE_2=ENABLE LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS1, PROD, DR) PROD SYNC to FS1 DR is Alternate LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS1, PROD, DR)' LOG_ARCHIVE_DEST_STATE_2=ALTERNATE LOG_ARCHIVE_DEST_3='SERVICE=FS1 SYNC AFFIRM MAX_FAILURE=1 ALTERNATE=LOG_ARCHIVE_DEST_2 DB_UNIQUE_NAME=FS1 VALID_FOR=(PRIMARY_ROLE,ONLINE_LOGFILE)' LOG_ARCHIVE_DEST_STATE_3=ENABLE 23
Add Second Far Sync Destination for HA Maintain Zero Data Loss if Far Sync Server Outage FS2 Ship to DR LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS2, PROD, DR)' LOG_ARCHIVE_DEST_2='SERVICE=DR DB_UNIQUE_NAME=DR VALID_FOR=(STANDBY_ROLE,STANDBY_LOGFILE)' LOG_ARCHIVE_DEST_STATE_2=ENABLE DR LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS1, FS2, PROD, DR)' PROD FS2 is alternate in place of DR LOG_ARCHIVE_CONFIG='DG_CONFIG=(FS1, FS2, PROD, DR)' LOG_ARCHIVE_DEST_2='SERVICE=FS2 SYNC AFFIRM DB_UNIQUE_NAME=FS2 VALID_FOR=(PRIMARY_ROLE,ONLINE_LOGFILE) LOG_ARCHIVE_DEST_STATE_2=ALTERNATE 24
Data Guard Broker and Far Sync Automatically configures ASYNC redo transport from PROD to FS1 DGMGRL> ADD FAR_SYNC_INSTANCE FS1 AS CONNECT IDENTIFIER is FS1; Manually configure DR to receive from FS1 using RedoRoutes DGMGRL> EDIT FAR_SYNC_INSTANCE FS1 SET PROPERTY RedoRoutes= (PROD:DR ASYNC) ; Manually configure primary to only ship to FS1 and modify to SYNC DGMGRL> EDIT DATABASE PROD SET PROPERTY RedoRoutes= (LOCAL:FS1 SYNC) ; 25
Distance Between Far Sync & Primary? Impossible to completely escape the impact of RTT network latency on performance For best protection, locate the Far Sync instance at the maximum distance an application can tolerate synchronous transport Different applications will have different tolerance to RTT latency New Data Guard 12c Fast Sync will increase the practical distance Near-zero performance impact if network RTT is less than local I/O time 26
Data Guard 12c Synchronous Transport Process Architecture same as Data Guard 11g Release 2 User Transactions Queries, Updates, DDL Commit Primary Online Redo Logs Queries Reports Testing Backups SGA Redo Buffer LGWR NSS Oracle Net RFS Standby Redo Logs MRP Primary Database Active Standby 27
Introducing Data Guard 12c Fast Sync Net Reduction in Round-Trip Time Improves Performance User Transactions Queries, Updates, DDL Commit Primary Online Redo Logs Queries Reports Testing Backups SGA Redo Buffer LGWR NSS Oracle Net RFS Standby Redo Logs MRP Primary Database LOG_ARCHIVE_DEST_3='SERVICE=stby1 SYNC NOAFFIRM' Active Standby 28
Data Guard 12c The Evolution Continues Production Offload, HA, Data Protection Far Sync zero data loss and production offload WAN zero data loss failover to remote standby Off-host redo transport compression requires ACO Gap resolution Serve up to 29 remote destinations Expand reporting offload to Active Standby DML to global temp tables Sequence support Use cascaded destination for up-to-date reporting Real-Time Cascade HA Upgrade - automation for database rolling upgrades Performance, Manageability, Feature Support Fast Sync Data Guard Broker Validate database - role change readiness Resumable switchover Automatic lag monitoring based on user configurable thresholds Support for cascaded standby and real-time cascade Configurable Broker tracing Support for Container Databases Separation of Duties - Data Guard Administrative Privilege 29
ASYNC transport mode
Data Guard Far Sync Putting an End to Compromise Zero Data Loss at any distance High Performance Simple Low Cost 41
Resources OTN Data Guard Portal: www.oracle.com/goto/dataguard OTN HA Portal: www.oracle.com/goto/availability Maximum Availability Architecture (MAA): www.oracle.com/goto/maa MAA Blogs: http://blogs.oracle.com/maa Oracle HA Customer Success Stories on OTN: www.oracle.com/technetwork/database/features/ha-casestudies-098033.html 42
Key HA Sessions and Demos by Oracle Development Monday, 1 October Moscone South 12:30p Oracle Data Guard Zero-Data-Loss Protection at Any Distance, 300 12:30p Future of Exadata: OLTP, Warehousing, and Consolidation, 104 1:45p Automating ILM with the Latest Database Technology, 300 1:45p Extracting Data in Oracle GoldenGate Integrated Capture Mode, 102 3:15p Maximize Availability with the Latest Database Technology, 303 3:15p Maximize Enterprise Availability with the Latest DB Technology, 303 4:45p Mission-Critical Oracle Exadata OLTP Deployment at PayPal, 300 4:45p Temporal Database Capabilities with the Latest DB Technology, 300 Tuesday, 2 October Moscone South 10:15a Database Tables to Storage Bits: Data Protection Best Practices, 300 10:15a GoldenGate & Data Guard: Working Together Seamlessly, 305 11:45a Active Data Guard Zero-Downtime Database Maintenance, 300 11:45a Using Automatic Storage Mgmt with the Latest DB Technology, 301 1:15p The Four Ts of RMAN: Tips, Tuning, Troubleshooting, and?, 102 5:00p Maximum Availability Architecture Best Practices for Exadata, 303 Demos Mon 10:00a-6:00p - Tue 9:45a-6:00p - Wed 9:45a-4:00p Oracle Maximum Availability Architecture, S-011 GoldenGate 11gR2: Real-Time, Transactional DB Replication, S-027 Oracle Database 12c: Global Data Services, S-010 Oracle Database 12c Application Continuity - S-009 Wednesday, 3 October Moscone South 10:15a Operational Best Practices for Oracle Exadata, 102 10:15a Maximize Availability by Minimizing Disruption for End Users and Application, 301 11:45a What s New in the Latest Generation of Oracle RAC, 301 11:45a Best Practices for HA w/ GoldenGate on Oracle Exadata, 102 1:15p Oracle Secure Backup: Integration Best Practices with Engineered Systems, 300 1:15p Application MAA Best Practices on Oracle Private Clouds, 200 5:00p Tuning &Troubleshooting Oracle GoldenGate on Oracle, 102 Thursday, 4 October Moscone South 11:15a Integrate Your Globally Distributed Databases for Key Cloud Computing Benefits, 300 12:45p Backup and Recovery of Oracle Exadata: Experiences and Best Practices, 300 Oracle Secure Backup, S-014 Oracle Active Data Guard, S-007 Oracle Recovery Manager and Oracle Flashback Technologies, S-019 Oracle Real Application Clusters and Oracle RAC One Node - S-008 Oracle Database 12c Xstream, Streams, Advanced Queing, S-018 43 After OpenWorld, visit oracle.com/goto/availability
The preceding is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle s products remains at the sole discretion of Oracle. 44
Graphic Section Divider 45
46