Agenda

1. Storage Edition for Oracle Storage Management Software Suite for Oracle Databases on HP ServersMelissa SteinProduct ManagementOctober 1999

2. Agenda • VERITAS Product Overview • DBA’s Challenges • VERITAS Storage Edition for Oracle • Enhanced Performance • High Availability • Faster Recoverability • Improved Manageability • Summary

3. Database Administrator’s Challenges • Optimal database performance • 24x7 availability of mission-critical databases • Continuous database access • Fast recovery • Restore from consistent backups • Increasing management complexity • High growth in data volume • Complex data types (e.g., video, sound) • Increasing number of users • Diverse database server configurations (e.g. replication servers)

4. Storage Edition for Oracle • VERITAS Storage Edition for Oracle enables database administrators to manage Oracle7 and Oracle8 environments with the flexibility of file systems at raw disk performance. • Its data redundancy techniques and online administration features increase data reliability and availability. Storage Edition for Oracle enables DBAs to meet their top priorities: optimal performance, availability, recoverability, and manageability of databases

5. Storage Edition 2.0 Architecture • Integration of online and offline storage management Oracle Storage Edition for Oracle VxDBA Utility NetBackup Block-level Incremental Backup Extension VERITAS File System with Quick I/O & Storage Checkpoints

6. Feature Overview • Quick I/O • Raw device performance with File System manageability • Asynchronous I/O support • Online administration of database storage • Storage Checkpoints • Enable Block-level Incremental Backup • Enable Storage Rollback for immediate “point-in-time” recovery from on-disk images • HP-UX 11.0 support

7. Storage Edition for Oracle Enhanced Performance

8. Storage Edition for OraclePerformance • Increased performance increased productivity • Equal to raw device performance using Quick I/O • Large I/O optimization • Accelerated I/O performance for large files (e.g., video, images)

9. Database Performance Issues with Conventional File Systems • Redundant data buffering • System buffer cache Database buffer cache • Redundant buffer copying • Inefficient use of memory • Single writer lock per file at UNIX file system level • UNIX file-level lock RDBMS row- or page-level lock • Serialized write operations • Impact on performance in multi-user client/server environments • Asynchronous I/O • Supported only on raw devices, not on regular files

10. Building Oracle Databases onRaw Disks vs. File Systems • Unique, total storage foundation solution to deliver the best of both worlds Storage Editionfor Oracle Raw Disks Conventional File Systems Advantages Disadvantages Advantages Disadvantages • Manageability • Tablespace • organization • System-wide • backup and • restore • policies • Easy to create • and expand • files for Oracle • Manageability • “Invisible” • locations for • tablespaces • Difficult to • grow • tablespaces • Performance • Lower database • throughput • (30-90% of • raw disk) • Higher CPU • overhead • Data Reliability • Potential data loss • Performance • Best OLTP performance • Data • Reliability • No potential for data loss with data written to • disks directly

11. Quick I/O Oracle Server raw device regular file Raw Disk PerformanceQuick I/O Database Accelerator • Break-through VERITAS File System interface technology • Presents regular VERITAS File System files to Oracle as raw character devices • Allows parallel updates to database files for increased throughput • Oracle handles locking for data integrity • Eliminates traditional UNIX file system overhead • Removes single-writer lock at file level • Bypasses redundant data buffering

12. Quick I/O PerformanceComparison to Raw I/O • Storage Edition has equal to raw performance • 160% performance improvement over traditional file systems • Tested configurationincluded: • HP K570 (4 CPUs and 1 GB of RAM) • HP-UX 11.0 (64-bit) • Storage Edition 2.0 for Oracle • Oracle 8.0.5 (64-bit) • TPC-C scale factor of 100 warehouses

13. Large I/O Optimization • VERITAS File System allocates disk space in groups of contiguous blocks or “extents” • Extent-based allocation can accelerate I/O by increasing the I/O transfer size in the File System • Optimal for large files • Large Oracle tablespaces • Binary Large Objects (BLOBs) (e.g., images, sounds)

14. Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Disk Block Veritas File Systm Inode Disk Block Disk Block Extent Addresses Disk Block Disk Block Direct Address 0 / Length Direct Address 1 / Length Direct Address 2 / Length Direct Address 3 / Length Direct Address 4 / Length Direct Address 5 / Length Direct Address 6 / Length Direct Address 7 / Length Direct Address 8 / Length Direct Address 9 / Length Single Indirect Block Double Indirect Block VERITAS File SystemExtent-based Allocation

15. Storage Edition for Oracle High Availability

16. The Need for High Availability • Many factors contribute to potential downtime • Software • Hardware • People • Planned downtime (for maintenance tasks) • Downtime impacts business operations • Revenue, customer service, and productivity at risk

17. Brokerage Operations Finance $6.45 Million Credit Card/Sales Auth. Finance $2.6 Million Pay-per-View Media $150 Thousand Retail Home Shopping (TV) $113 Thousand Retail Catalog Sales $90 Thousand Transportation $89.5 Thousand Airline Reservations Media $69 Thousand Tele-Ticket Sales $28 Thousand Package Shipping Transporation Finance $14.5 Thousand ATM Fees Financial Impact of System Failure Application Industry Average Hourly Cost Source: Contingency Planning Research

18. Storage Edition for OracleAvailability • Online administration • Storage reconfiguration and maintenance without downtime • Automated, proactive monitoring of file system to avoid unexpected “out-of-space” conditions

19. Online Administration • Enables DBAs and system administrators to meet changing database storage requirements without database downtime • Supports online tablespace growth • Online resizing of file systems • Allows database storage reconfiguration • Defragmenting mounted file systems

20. Extend a file system • Locate new disk space • Create a new file system • Stop work • Copy files to the file system • Resume work • Select “resize a file system” from a menu • Enter the new file system size • No interruption to users or applications Online Administration Example

21. REACTIVE PROACTIVE Intelligent File System Control • Policy driven File Management • Why ? • Lack of file system free space can cause widespread application failure • How ? • Included Agent monitors average available free space of VERITAS File System • If average free space falls below threshold set by the administrator, the agent will automatically grow the File System and underlying volume if need be to enforce free space policy • Configured via VxDBA utility

22. Storage Edition for Oracle Faster Recoverability

23. Storage Edition for OracleRecoverability • Fast database recovery from system crash • Fast file system recovery through journaling • Database recovery from consistent, stable backups • File System snapshots for system and database backups • Virtually eliminate database backup window • Enables block-level incremental backup to back up only changed data blocks • Faster database Recovery from logical error • Enables point-in-time recovery from on-disk image

24. Database Recovery • Database recovery begins only after file system recovery • Journaling File System minimizes database downtime • File system recovery only requires replaying the intent log • File system recovery time does not depend on file system size • Large file system (over 100 GB) usually takes only seconds to verify / recover

25. Consistent Database Backups • VERITAS File System’s snapshot mechanism • Enables system-wide backups for database and non-database files • Reduces backup window for cold database backups to a couple of seconds • Simplifies recovery process for hot database backups • Supports UNIX utilities like tar and cpio, and off-the-shelf UNIX third party solutions

26. VERITAS File System Snapshot Take a snapshot at 9 AM on Monday /oradata_snapshot /oradata • “Original” data blocks • A consistent “view” of • the file system at 9AM • by keeping track of • changed data blocks • (“copy on write”) • Only 5-20% of original • storage required Data blocks continue to be updated

27. Shutdown the database A couple of seconds of downtime Snapshot Start the database Mount snapshot file system(s) Perform backup Resume operation Unmount snapshot file system(s) Full Database Backups withFile System Snapshot • Cold database backup (or off-line backup) • Enables database access and backups in parallel • Small storage requirement (approx. 15% additional disk space)

28. Begin tablespace backup A couple of seconds of changes to apply at recovery Snapshot End tablespace backup Mount snapshot file system(s) Perform backup Hot Database Backups withFile System Snapshot • Hot database backup (or on-line backup) • Reduces recovery window with fewer changes to apply

29. Key Differentiator:VERITAS Storage Checkpoints • Point-in-time snapshot of database • Storage Checkpoints maintain changed data blocks as database is being updated • Uses efficient copy-on-write technique • No post-processing needed to scan forchanged data blocks

30. A B C D E Create a Checkpoint.. Simplified illustration of initializing a storage checkpoint Primary File System Storage Checkpoint (empty) The storage checkpoint presents an exact image of the primary file system.

31. A B C D E and update the filesystem Simplified illustration of initializing a storage checkpoint Primary File System Storage Checkpoint Update E1 E As the primary file system is updated, the original data is copied to the storage checkpoint before the new data is written. As the primary file system continues to change, the storage checkpoint accumulates the original data blocks

32. A B C D E and again.. Simplified illustration of initializing a storage checkpoint Primary File System Storage Checkpoint Update C C1 E1 E And so the process continues until the next checkpoint is created

33. Benefits of Storage Checkpoints • Block-Level Incremental Backup: • NetBackup manages Storage Checkpoints and extracts only the changed blocks • Storage Rollback: • VxDBA utility manages Storage Checkpoints for faster database recoverability from on-disk backup image

34. Traditional Backup Approaches • Full backups • Shrinking backup windows • Approaches to minimize downtime • Hot database backups • Third mirror break-off • Rolling backups (full + tables) • File-level incremental backups • Poor fit for database environments • Other (block-level) incremental backups • SQL-BackTrack • Oracle8.0’s Recovery Manager (RMAN) • Our advantage: Storage Checkpoints

35. Storage Edition:Block-Level Incremental Backups • Back up only changed data blocks! • Higher data availability • Virtually eliminates backup windows • Allows more frequent backup schedules • Backups contain up-to-date data • Enhanced performance • Significantly lowers CPU and network resource drain during backups • Faster recovery from backups (less redo logs to apply) • Requires NetBackup Server engine 3.2 and and database server to be backed up to be a NetBackup client

36. Block-Level Incremental Backup NetBackup Server Changed BlocksSince LastBackup VERITAS File System with Storage Checkpoints NetBackup BLI Extension Minimize Your Database Backup Window BlockLevelIncrementals PreviousFull Backup FullRestore +

37. Cold Database Backupswith Storage Checkpoints • Enables database access and backup in parallel with minimal interruption • Cold database backup (or off-line backup) with a couple of seconds of downtime Shutdown database A couple of seconds of downtime Storage Checkpoint Perform backup Start the database Resume operation Managed by NetBackup

38. Hot Database Backupswith Storage Checkpoints • Hot database backup (or on-line backup) • Reduces recovery window with fewer changes to apply Begin tablespace backup A couple of seconds of changes to apply at recovery Storage Checkpoint Perform backup End tablespace backup Managed by NetBackup

39. storage checkpoint at 1:00AM /oradata • Bring the database into a consistent state A • Take a storage checkpoint • Resume database operation • Back up from the storage checkpoint • The storage checkpoint presents a “frozen” image of the file system B C D E F G H Full Backup with Storage Checkpoint A B C

40. /oradata A B C D E F G H A B C During Full Backup storage checkpoint at 1:00AM • During backup, the database is accessible and being updated • The storage checkpoint continues to present a “frozen” image of the file system A’ A G’ G D E F G H

41. storage checkpoint at 1:00AM /oradata A B C D E F G D’ G’ H’ A’ H After Full Backup • After the full backup, the database continues to be updated A’ A D’ D So… Which blocks have changed since the full backup? What are the changes? G’ G H’ H

42. Second storage checkpoint at 12:00PM First storage checkpoint at 1:00AM /oradata A’ A B C D’ D E F G’ G H’ H Block-Level Incremental Backup • Bring the database into a consistent state • Take a second storage chkpt • Resume database operation • Back up using the storage checkpoints • Before-images saved on the first storage checkpoint indicate which blocks have changed since the full backup • Changed data is read from /oradata D’ A’

43. Second storage checkpoint at 12:00PM First storage checkpoint at 1:00AM /oradata A’ A B C D’ D E F G’ G H’ H D’ A’ During Block-Level Incremental Backup • During backup, the database is being updated • Changes after 12:00PM are saved on the second storage checkpoint • The second storage checkpoint presents a “frozen” image of /oradata at 12:00PM • Changed data is read from the second storage checkpoint or /oradata C C’ D* D’ H’ H* G’ H’

44. Restore from BLI Backup • Restoring incremental image requires restoring full backup plus all incrementals to get to desired point in time. • Benefit: BLI Backup enables more frequent backups • More up-to-date backup images • Less logs to apply upon restore • Faster database recovery BlockLevelIncrementals PreviousFull Backup FullRestore +

45. Storage Rollback • Storage Checkpoints are on-disk backup images • Consistent, stable images of a file system at a particular point in time • Storage Rollback is “restore” from the on-disk backup images or Storage Checkpoints • Copies before-images of data blocks back to the live file system • Restores without going to tapes

46. Storage Rollback Recovery from Storage Checkpoints = Simply write back changed (blue) blocks to previous state and then apply redo logs to bring database to desired point in time

47. Storage Rollback Example Second storage checkpoint at 12:00PM First storage checkpoint at 1:00AM /oradata Images of /oradata and its Storage Checkpoints A’ A B C’ C D* D’ D E F G’ G H* H’ H

48. Storage Rollback Example /oradata at 1:00AM /oradata at 12:00 PM A A’ B B C C D D’ E E F F G G’ H H’

49. C D’ H’ Storage Rollback to 12:00PM Second storage checkpoint at 12:00PM /oradata /oradata at 12:00 PM A’ A’ Before-images on the 12:00PM Storage Checkpoint are copied back to the live file system, /oradata B B C C’ C D’ D* D’ E E F F G’ G’ H’ H* H’

50. Storage Rollback to 1:00 AM Second storage checkpoint at 12:00PM First storage checkpoint at 1:00AM /oradata at 1:00AM /oradata A A’ A A Before-images from 1:00AM and 12:00PM Storage Checkpoints are copied. B B C C C’ C D D D* D’ D E E F F G G’ G G H H H* H’ H