Closed eljhkrr closed 1 year ago
If we delete the data, it means we can't reuse the informaiton already available for other purposes. If you want to ensure that reporting is not impacted because of the huge dataset, we can drop the old data in historical tables and keep reporting data for 6 months only. Here's how we are doing that for impact monitoring.
This way, we can keep the scraped data for longer term, yet keep current reporting data small and fast.
Thanks @yrimal, added as agenda item for upcoming meeting
If the target table is well formed, you wouldn't have an issue with a proper indexing strategy
This is how many docs we have. We have been monitoring for about a year. It is about 50MB of data. Should we just ignore this issue for now?
doctype | count |
---|---|
klipfolio_datasource | 8,944 |
monitoring | 12,127 |
error | 113 |
couch2pg | ~5k |
Should we just ignore this issue for now?
I agree, this sounds like a nominal data.
Proposed period for maintaining records is 6 months as we're unlikely to act on older data