Case Study – Midwest Electricity Provider
“After installing an ERP application, there was no way we could manually handle the large number of objects in our database. CDB/Automation made it possible to dynamically control execution of our utilities at critical points.”
Environment
- Over 200 gigabytes of production data in DB2
- Largest application contained information for 5.5 million customer accounts
- New ERP application introduced huge number of new objects
- Need to meet Service Level Agreements (SLAs) for application availability and offsite recovery
- 24x7 availability needed for Plant operations
Implementation
- Used Wildcards to process many objects per job
- Reduced number of jobs
- Tolerates changes in environment (i.e., dropped or added objects)
- Included rules in exit to tailor processing to individual applications
- Some require dual, offsite copies
- Some applications run constantly (e.g., plant applications)
- Some applications run less frequently (e.g., billing, monthly reporting)
- Chose to Image Copy every object processed. Automation determined:
- Full or incremental copy
- Single or dual copy
- Real or Virtual tape
- Which copies to stack together on tape
- How many tape drives to use (based on availability at run time)
- Whether or not to cleanup SYSCOPY
- Chose to Reorg only objects which met criteria defined on an application basis, such as:
- Clusterratio threshold
- Number of extents
- Day of month
- Index only or Tablespace Reorg required
- Used Automation to automatically size and allocate datasets
Benefits Realized
- Met all Service Level Agreements
- Reduced workload of DB2 Database Administration staff
- Fewer jobs to maintain
- Fewer job failures due to human error
- Eliminated job failures due to insufficient allocations
- No updates required for new or dropped objects
- New applications seamlessly integrated
- Eliminated manual cleanup of SYSCOPY
- Reduced System Resources
- Eliminated unnecessary utility runs resulting in CPU and Elapsed time savings
- Correct size allocation of datasets resulting in DASD savings
- Processed Reorg and Copy in one pass of the tablespace for CPU/Elapsed time savings
- Better and more consistent application performance
- Objects reorged as often as necessary
- New applications included in utility processing immediately
|
