Optimization of TCO - part II: Run

As data volumes grew, an already full run schedule became a puzzle and a hot topic: when to run what to ensure timely data delivery, whilst not impeding other activities such as sofware deployment, remedial actions and new functionality.

First, analysis took place to find the poor performers. These were further investigated to identify the main causes of the long running jobs and views. Summarized, not enough or late filtering, locking of main used tables, concept core views whose execution only grow with time, core conceptual approach on how to load non-temporal artifacts (e.g. balances) were the main culprits. A prioritized plan was created to deal with the various causes, not with a band-aid but in a way that is structural. Changing concepts leading to leaner processing, reducing TCO in short, mid and long term. It's about doing the right thing and doing things right.