Hi BWA Experts,
We need to draft a policy or process on when to exclude cubes/queries from BWA. I know problematic or rogue queries are hard to identify unless there's BWA statistics or an ongoing python trace ongoing. I would like to ask the community if they have implemented such policy as well and how? For example:
1. Is BWA statistics by default turned on? Does the support group check or analyze data on a daily/weekly basis? What is being looked at for these statistics to identify the "bad" queries?
2. If not BWA statistics, are there other tools?
3. Once a cube or query is identified to be causing issues, what testing can be done in order to deem these objects as "fit" to be brought back to BWA?
Hoping for a favorable response. Thank you.
Regards,
Nikos