README.md

etcd defragmantation cronjob

This is a simple cronjob checks whether the etcdDatabaseHighFragmentationRatio alert is active for the cluster or not; if it is active, then it runs etcdctl defrag command for each etcd instances starting from the ones that are not leader.

Some Notes

• Version of etcdctl is 3.5.3
• Prometheus service and its namespace should be set as an environment variable on the cronjob.yml.
• A secret object in which the etcd client certs and CA should exist.
• An endpoint object consisting of etcd endpoints.

References and Useful Discussions

• etcd-io/etcd#14975
• https://etcd.io/docs/v3.5/op-guide/maintenance/#defragmentation
• https://www.gojek.io/blog/a-few-notes-on-etcd-maintenance
• https://coderise.io/etcd-on-kubernetes-with-high-availability-maintenance-part-3/
• https://kubernetes.slack.com/archives/C3HD8ARJ5/p1661525680896279