Facilitate Failover Test of Staging Using Geo

As part of the deliverables of the GitLab.com Disaster Recovery Working Group we would like to conduct a test of a Geo-enabled failover for staging.gitlab.com.

This test would be a planned failover of staging at a scheduled time, which:

1. Fails over staging.gitlab.com to a Geo-based secondary
2. Tests that staging.gitlab.com works on this secondary
3. Fails back over to the original staging.gitlab.com infrastructure.

What needs to happen to facilitate this:

• Review the current setup of Geo in staging with the Geo team.
• Determine any blockers to a staging failover and work with the Geo team to get them addressed.
• Build a change management issue to track the tactical pieces to this failover.
• Coordinate with various consumers of staging to help define a testing schedule and what we will need to have in place in the event we are without staging for an extended period of time