© 2006 ibm corporation maintaining high database performance ims db recovery solution the ibm ims...
TRANSCRIPT
© 2006 IBM Corporation
Maintaining High Database Performance
IMS DB Recovery Solution
The IBM IMS Tools Solution
Margaret M WilsonIMS Tools Market Manager
2
Highlights
We understand your company’s needs…and we’re here to help
The product families we offer
– Systems Administration
– Recovery
– Reorganization
– Performance and Tuning
– Fast Path Solutions
– Data Administration
Summary
3
Solutions that fit your needs
IBM IMS Tools
Systems Administration
Recovery
Reorganization
Data Administration
Fast Path Solutions
Performance and Tuning
4
Recovery Solutions
IMS High Performance Image Copy
IMS High Performance Change Accumulation Utility
IMS Database Recovery Facility
IMS Index Builder
5
A reliable recovery plan is critical to your business
Your mainframe databases represent a huge investment by your company.
If your data is unavailable for any reason, you are not receiving the expected gain on your investment.
When this occurs, IT becomes a cost burden instead of a business asset.
An e-business must be available and operational 24 hours a day, 365 days a year.
6
Some database outages are planned…
Application database maintenance
Data migration
Database design change implementation
Hardware maintenance or upgrades
Disaster recovery preparation and testing
7
But the highest impact outages are not… Application errors
Improper data feeds
User errors
Operations errors
Batch cycle errors
Fallback from application change migrations
Site disasters (floods, power outages, storms, fires and so on)
Hardware failures
8
IBM IMS Tools can help
When these situations occur, a recovery solution that is complicated and time consuming to prepare and execute will result in increased down time and loss of revenue.
Software based recovery tools are not a luxury or an insurance policy — they are a key component of mission
critical database support.
9
The challenges with using IMS base utilities for recovery
IMS recovery is a very complex task, quite prone to errors and generally difficult to manage.
Utilities use serial processing adding to the time your data is unavailable.
Less automation means more manual labor and more potential for error.
All the while your databases are offline and unavailable for your critical business needs.
10
A hardware solution isn’t enough Hardware solutions are not complete.
Hardware only covers what was mirrored.
It’s too expensive to send all of your business critical databases offsite.
Even with a hardware solution, recovery tasks are still needed to get back up and running.
What about application program errors?
Having multiple copies of logically incorrect data is not a solution.
11
The IBM IMS Tools Recovery Solution
The IBM IMS Tools Recovery Solution is an effective and efficient way to ensure your IMS data is recovered quickly and accurately.
These four tools combine to form an integrated and automated solution to meet your recovery needs
– IMS High Performance Image Copy
– IMS High Performance Change Accumulation Utility
– IMS Database Recovery Facility
– IMS Index Builder
12
IMS High Performance Image Copy v4.1 5655-N45
Creates exact (image) copies of IMS databases
Functions:– Run an Image Copy function with the hash checking of IMS HP Pointer
Checker V2 under the control of IMS Parallel Reorganization, V3• Major reduction in run-time of the reorganization process • Supports processes in parallel within an IMS Parallel Reorganization
– New optional parameter for controlling the secondary image copy data set – Performs accuracy checking of an image copy– Dynamic Allocation for all input and output datasets– Reports DBRC information about the registered Image Copy data sets
Your Value:– Reduces operational costs by using less tape storage– Minimizes manual steps involved with tape usage– Automated functions help data centers run more efficiently
13
IMS High Performance Image Copy v4.1 5655-N45 Advanced copy services
– Concurrent copy,
– FlashCopy and
– SnapShot copy.
Reduces image copy and recovery time, reduces elapsed time and CPU utilization.
– Minimizes CPU utilization and elapsed time by using High Performance (HP) Input/Output function for read and write processing
Provides automatic checkpoint and restart. – Restart failed image copy jobs by resubmitting the same job stream.
– No alterations to the job stream are necessary, which avoids duplicating previously successful image copies.
Stops and starts database automatically.
– IMS Tools Online System Interface
14
IMS HP Change Accumulation v1.3 5655-F59 Runs multiple change accumulation groups in parallel and streams output
across all addresses simultaneously
Functions:– Provides both ISPF interfaces that allow for customization, setup and execution and
Batch interfaces for operations– Elimination of separately scheduled Fast Path sort address space– DBRC is now optional– Processes input data (from RLDS's) and output data (new CA's) in parallel
• Users can control the number of parallel processes – Improved reporting
• Consolidates the status of associated address spaces into a single report, and extends diagnostic capabilities with new error reporting messages
Your Value:– Increased database availability– Increased IMS database data integrity
Control SpaceControl Space
Main TCB
LogReader
TCB
IMSLog
IMSLog
LogReader
TCB
SORT SpacesSORT Spaces
Main HPCA TCB
SORTProduct
TCB
E15 E35
NewCA
OldCA
15
IMS Index Builder v2.3 5655-E24
Helps streamline index maintenance and recovery
Functions:
– Add or Rebuild all or some secondary indexes of an IMS full-function database (including HALDB)
– Build both primary and secondary indexes in one job step, eliminating the need to image copy indexes
– Rebuild HIDAM or PHIDAM partition primary index
– DBRC notification support
– Provides an ISPF front end for easy JCL generation
– Creates separate sort address spaces
– Duplicate Record Processing to Spill File
Your Value:
– Improves overall performance
– Improves DBA productivity by its’ enhanced ease of use
– Saves time by performing index recovery without full database reorganization by using DL/I to scan the databases
More information
16
IMS Database Recovery Facility v3.1 5655-N47 High Performance IMS Database Recovery Solution - Recovers IMS
database data sets and Fast Path areas Functions:
– Recovery of Multiple DBDS or Areas in one step
• Runs in Parallel with IMS Transaction Processing• Online or Batch mode recovery
– Supports all Recoverable IMS Database Types
• Not GSAM, HSAM, SHSAM, or MSDB types– Full formatted recovery report
– Report only execution (VERIFY option)
Your Value:
– Increased database availability:
• Applying database changes to multiple database data sets simultaneously with a single pass of input data
• Reducing recovery time
– Enhanced usability and flexibility by supporting report-only execution
17
IMS Database Recovery Facility v3.1 5655-I44
Increased Tools Integration
– Automatic invocation of High Performance tooling, including Image Copy, Pointer Checker, Fastpath Pointer Checker, Index Builder, and DFSPREC0.
Database Copy Generation
– Creation of copies of database data sets without accessing production copies
– Input is image copy, change accumulation, and log data sets
– Copies can be created to any point in time via PITR
18
IMS Database Recovery Facility v3.1 5655-I44
Automatic Delete/Define of Database Data Sets
– Output data sets are created as part of the recovery process
Incremental Image Copy
– Created via HPIC
Allocate/Open Option on START VERIFY
– Logical and physical validation of data set availability prior to running the actual recovery job
Submitting IMS Commands via TOI/XCF
– DBR local or global
– START local or global
19
IMS DRF v3.1 5655-I44 – Recovery Flow
Use dynamic allocation of data sets– database, log, change accumulation, and image copy
Does NOT use IMS resources for recovery
Gets all information about recovery datasets from RECONs– The Catalog is used if the RECON has not recorded data set migration
Read Log data sets in parallel– From RLDS or SLDS only– Single Log Pass for Multiple DBDS Recovery
Read Change Accumulation datasets concurrently with Log data sets– Complete or Incomplete (with Spill records)– Change Accum is not required, including in a data sharing environment
Sort Log Records
Update Image Copy data with log and change accum data in a single pass– Standard IC - Changes applied as data is read– IC2 - Data is restored - Only updated records are read
20
The IBM IMS Tools Recovery Solution
Start OptionsIMSIMS
JCLJCL
DRFDRF
Recon
Master AS
Subordinate ASControl Function
DBRC
Log and CA read
Log
IC DBDS
AreaData Set Restore
SortSub ASSub AS
Data Set Restore
Sort
Report
IB/DFSPREC0 AS
IB or DFSPREC0
HPPC, DEDB PC HPIC
IC
Recovery Flow
Build primary and secondary index data sets during the recovery process
Rebuild post recovery phase
HALDB Primary Index and Index List Data Set rebuild
Rebuild post recovery phase
Hash check pointer validation run as part of recovery processing
Block by block pointer check, not post processed
Produce image copies as part of recovery processing
Image copies are registered with DBRC
Block by block image copies
21
Typical recovery scenarios
Application Error
Producing An Audit Copy
Disaster Recovery
1
2
3
22
Application Error
Point-In-Time Recovery
– Need to determine what application and when
– Use IMS Tools Recovery Solution to recover the effected databases
• Recover all databases in parallel• Read multiple logs in one pass• Recover only committed data to the time specified• Recovery happens without backout
1
Time
Committed data
Non committed data
Run away application submits update twice
23
Producing An Audit Copy
Creating an audit (or test) copy of your databases using the IMS Tools Recovery Solution
– Produce copies without accessing production databases
– Build copies from image copy, change accumulation, and log data sets
– Generate data set names• Prefix or suffix• Date, time• High level qualifier• Literal
2
DRF
IMS Production
Database
CA, Log, IC
Database Copy
24
Disaster Recovery
Moving your data from a remote site to your production site in case of a disaster
– Option 1
• Send a copy of your image copies and logs offsite• Then ship them back to your production site when needed• The IMS Tools Recovery Solution will use them as input to
rebuild your databases
– Option 2
• Perform change accumulation using HPCA• Use this as input to make an incremental image copy using
DRF and ship offsite• When shipped back to your production site, DRF recovers
to an image copy
3
25
Summary
The IMS Tools Recovery solution combines the power of four high performance tools from IBM to help you quickly and efficiently recover your IMS data…
Supports a variety of image copy formats as input including concurrent copy, FlashCopy and SnapShot copy and leverages High Performance Input/Output functions.
Quickly and efficiently reads multiple logs in parallel, processes that data, and sends it to multiple address spaces for multiple CA groups.
The cornerstone of the IMS Tools Recovery Solution, integrating the power of Image Copy, Change Accumulation, and Index Builder tooling under one interface.
Eliminates the need for recoverable indexes, which eliminates the need to image copy indexes and reduces IMS log volume.