GitLab.com scale
In this video, John Northrup from our infrastructure team discusses HA in a fireside chat on how we've configured gitlab.com's HA/Geo set up. gitlab.com supports thousands of users and our customers often ask how we've scaled this and what would be ideal for them to set up themselves.
Topics covered:
- Failover state
- Monitoring
- What portions of the GL setup are stateless and which are stateful?
- NFS Redundancies
- Sidekiq Queue Splitting
- PG Bouncer
- Redis Sentinel
- GitLab Service Discovery
- HA Proxy
- Cloud vs. on-prem
Video
Next
You can close this issue now