monogon
monogon copied to clipboard
Metropolis Control Plane HA
We've never really tested Control Plane HA in an end-to-end scenario, and we reaped the effects of that during our first large production deploying.
Related fixes:
- https://review.monogon.dev/c/monogon/+/2067
- https://review.monogon.dev/c/monogon/+/2068
- https://review.monogon.dev/c/monogon/+/2069
- https://review.monogon.dev/c/monogon/+/2071
E2E tests are still in progress.
Forensics data from a recent HA control plane failure:
https://drive.google.com/drive/folders/1knvTErYqXKEXUyCxvaNpHyM4tikz7QXY
Our assumption is that this fixes that, right? https://review.monogon.dev/c/monogon/+/2873
Closing unless proven otherwise - discussed with Lorenz and we're not aware of any outstanding issues (beyond more QA burn-in time).