Automation has to be central to everyone’s management of Db2. Without automation, there is just far too much to do, in far too little time, by too few people. Come and join me for “Surviving or Thriving – Automation may be the key to your future success” at IDUG North America in Charlotte, NC (from June 2nd to 6th, 2019) which walks us through the history of automation and stresses why choosing the right automation strategy is key.
The Birth of Automation
In the early days of Db2, many of the housekeeping tasks (such as reorganisations and taking backups of the data) lent themselves to simplistic automation. This was purely a JCL generation exercise when DBAs began to realise that the JCL they were executing was significantly repetitive and lent itself to creation by automated routines.
At first, JCL was created statically and unconditionally, based on the object names found in the Db2 catalog. Writing a simple program to query the catalog and write out JCL was both trivial and a significant step forward. No longer would newly-created objects be inadvertently skipped, nor would dropped objects cause processes to fail – risks that are inherent in all statically created jobs.
The next step, mainly confined to generation of Reorg JCL, was to make the creation CONDITIONAL – based on other catalog values, such as the data disorganization statistics. In this way, reorgs could be created, scheduled and executed only when they were actually needed – and an object that did not need to be processed could be skipped. When considering reorgs as a tool to enhance performance, it is important to focus on where that delivers results, and not to waste time and resources where no improvement can be expected.
Pros and Cons
This simple automation took a considerable load off the shoulders of the DBAs, but as time went on it became more complex. As the number of objects increased and the selection criteria grew, it became clear that home-grown solutions were not sustainable. Maintaining the automation itself became a significant drain on staff time – just the opposite of what it was supposed to be doing.
Additionally, collecting the statistics needed to drive the conditional creation was not free. Statistics collection comes with costs (both CPU costs and application availability conflicts). These costs only continue to rise as Db2 tables/indexes become ever larger.
There was also the worry about the gap between generation of some JCL and its execution. This could range from a few minutes to the best part of a week. If a possible improvement has been identified, surely action should be taken immediately and not deferred to a later time.
A better solution was needed. Both to manage the execution (or not) of activities and the timeliness of those executions.
Back to the Future
So, please join me at session B01 at IDUG (on Monday June 3rd at 12:45pm, currently planned for the Carolina C room) and discover the automation advantages that can help turbocharge your digital transformation and help you to increase your business agility. You will leave armed with next steps for implementing a database automation strategy that will:
- Help increase your business’ revenue and competitiveness
- Ensure stability and cost optimization
- Divert valuable staff time to more pressing needs, such as SQL statement tuning
These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.
See an error or have a suggestion? Please let us know by emailing firstname.lastname@example.org.