Spike: investigate GCP Managed Prometheus
Summary
From a Slack thread with GCP: https://gitlab.slack.com/archives/C01KPV0V3SM/p1646241445623149, Managed Prometheus is now GA https://cloud.google.com/stackdriver/docs/managed-prometheus
Related Incident(s)
None
Originating issue(s): gitlab-com/gl-infra/production#ISSUE_ID
Desired Outcome/Acceptance Criteria
- Investigate our Prometheus data volume / label space and make sure GCP managed will scale to our current load.
- How would this integrate with Thanos?
- Define what a viable POC would be.