etcd db cleanup maintenance and automation without downtime

If you have bootstrapped a etcd cluster and seeing etcd database space exceeded problem like the one below then this blogs helps you to solve :)

Error seems familiar and painful :)

etcdserver: mvcc: database space exceeded

The default storage size limit is 2GB. So etcd cluster needs periodic maintenance to remain stable and consistent.

All etcd storage resources consumed by the etcd keyspace. If an etcd member runs low on space, cluster-wide alarms which will put the system into a limited-operation maintenance mode.

To avoid running out of space the etcd keyspace history must be compacted.

We will see how this maintenance can be automated and performed without downtime

For #1 — If using the helm charts just enable the following values in your values.yaml

For #2 — here is the code for your K8s cron which take care of running defragemention.

Related posts:

The rhythm of a soulless night Follows along a path of greatness The glory you’ll succumb to If all else fails Don’t wish for love or pain This path will lead you nowhere Just follow along the dirt…

All of the incredible features you’ve read about the iPhone 14 are available on the iPhone 14 Plus. You won’t be missing out on anything by choosing the iPhone 14 Plus over the iPhone 14. When it…

Great Horned Musing. “Owl” is published by Rusty Alderson.