📹 Video Demo 📹
- Create 2 clusters (east, central)
- Add kubeconfig contexts for both
- Enable ncesaary GCP API's
gke-msc.sh
- Check MCS membership list
- Deploy Vitess operator to both clusters
- Create EtcdLockserver and ServiceExport spec in both clusters
- Wait ~5m for service imports to become available in both clusters
- Deploy Vitess cluster in central
- Deploy Vitess cluster in east
- Check logs for cross-region connectivity errors
- Confirm pod communication with orchestrator
./orcpf.sh
andhttp://localhost:3000
- Confirm reads/writes in both GKE clusters
- DONE!
- Setup Multi-Cluster-Service in GKE
- PlanetScaleDB Operator API
- Manifests and Config
- Vitess Operator
- Vitess Operator Federation, Architecture and Requirements
- 4/24/21: Added support for distributed Global Lockserver. Now the first 2 etcd instances are deployed to central while the 3rd is deployed to east. This has been tested and also working with 3 clusters (each having a single Global lockserver etcd instance)