As a result of https://issues.redhat.com/browse/MON-1563, we brought our findings to the Prometheus storage working group in order to get feedback. From this discussion, a couple of issues needed to be opened upstream. Since some of them have not been created, we need to follow up on that and make sure they are created and prioritized inside of our group as part of the Resiliency epic.

Conclusions of the storage working group discussion from the 05/18 can be found in this doc.

DoD:

• Follow up on the following issue and prioritize them:
  • WAL replay improvement: https://github.com/prometheus/prometheus/issues/6934
  • Move Index to external memory - Follow up with brancz on that one since he has the knowledge and context
• Create an issue and drive:
  • Early stale chunk flush