CKA — etcd Best Practices

Backup Frequency

Environment	Recommended Frequency
Production (critical)	Every 1-2 hours
Production (standard)	Every 4-6 hours
Staging	Daily
Development	Before changes
CKA exam	Before upgrades or destructive operations

Automated Backups

Create a cron job for automated etcd backups:

# Example: Backup every 6 hours
0 */6 * * * ETCDCTL_API=3 etcdctl \
  --endpoints=https://127.0.0.1:2379 \
  --cacert=/etc/kubernetes/pki/etcd/ca.crt \
  --cert=/etc/kubernetes/pki/etcd/server.crt \
  --key=/etc/kubernetes/pki/etcd/server.key \
  snapshot save /backup/etcd/snapshot-$(date +%Y%m%d-%H%M).db

etcd Performance

Factor	Recommendation
Disk	Use SSD storage for fast write operations
Network	Low-latency network between etcd members
Memory	etcd should have dedicated memory (not shared with heavy workloads)
Compaction	etcd auto-compacts by default; monitor disk usage
Defragmentation	Run periodic defragmentation to reclaim disk space

Check etcd Disk Usage

du -sh /var/lib/etcd/

Separate etcd for HA

In large production clusters, consider running etcd on dedicated nodes (external etcd topology):

Topology	etcd Location	Pros	Cons
Stacked	On control plane nodes	Simpler setup, fewer machines	etcd shares resources
External	On dedicated nodes	Better performance, isolation	More infrastructure

Disaster Recovery Planning

1. Regular backups to offsite storage (S3, GCS)
2. Test restores periodically in a staging environment
3. Document the restore process for your team
4. Multi-region backups for critical clusters
5. Encrypt backups at rest (etcd data contains Secrets)

Exam tip: While automated backups and HA are not directly tested, understanding these concepts helps with troubleshooting questions about etcd health and data recovery.

---

Next: Hands-On Lab ->