Etcd-backup-restore is a collection of components for backing up and restoring an etcd cluster. It also validates the etcd's data directory and auto-triggers restoration when needed.
etcd-backup-restore currently provides the following capabilities (the list is not comprehensive):
- Backups: etcd database snapshots (a.k.a full snapshots) and incremental snapshots are taken regularly, compressed and stored in the configured object storage provider.
- Data directory validation: It validates the etcd data directory before starting any etcd member.
- Restoration: In case of a database corruption for a single-member cluster it restores from latest set of snapshots (full & delta).
- Zero-downtime member recovery: Replaces a failed member in a multi-member etcd cluster without any downtime.
- Automated maintenance: Auto runs etcd's database maintenance operations such as defragmentation on a configured schedule.
- Secondary backups: In addition to primary backup storage, it uploads backups to a secondary object store, if configured, for higher backup resilience.
- Metrics: Exposes metrics to monitor the health and status of etcd-backup-restore operations.
Please check the Getting started guide for instructions on installation and basic usage. For detailed documentation, head over to the docs.
If you wish to contribute then please see our contributor guide.
We always look forward to active community engagement. Please report bugs or suggestions on how we can enhance etcd-backup-restore on GitHub Issues.
This project is maintained under Apache-2.0 license.