etcd backup and recovery in Kubernetes - Practice Problems & Coding Challenges

Challenge - 5 Problems

💻 Command Output

intermediate

Identify the output of an etcd snapshot save command

What is the expected output when you run the following command to save an etcd snapshot?

etcdctl snapshot save /tmp/etcd-backup.db

Kubernetes

etcdctl snapshot save /tmp/etcd-backup.db

A Snapshot failed: permission denied

B Error: snapshot save requires --endpoints flag

C etcdctl: command not found

D Snapshot saved at /tmp/etcd-backup.db

💡 Hint

The command saves the snapshot file to the specified path if permissions and environment are correct.

The 'etcdctl snapshot save' command saves the current etcd data to the specified file path. If successful, it outputs a confirmation message with the file path.

🧠 Conceptual

intermediate

Understanding etcd snapshot restore behavior

When restoring an etcd cluster from a snapshot, which of the following is true about the restored data directory?

A The restored data directory must be empty or non-existent before restore

B The restore command appends data to the existing data directory

C Restore overwrites only the WAL files but keeps existing snapshots

D Restore merges snapshot data with current cluster data automatically

💡 Hint

Think about how restoring a snapshot replaces the cluster state.

The etcd snapshot restore command requires the target data directory to be empty or not exist because it creates a fresh data directory from the snapshot. It does not merge or append data.

❓ Troubleshoot

advanced

Troubleshoot etcd snapshot restore failure due to permission error

You run the command:

etcdctl snapshot restore /backup/etcd-snapshot.db --data-dir /var/lib/etcd-new

and get the error:

restore failed: open /var/lib/etcd-new/member/snap/db: permission denied

What is the most likely cause?

A The etcdctl version is incompatible with the snapshot version

B The snapshot file /backup/etcd-snapshot.db is corrupted

C The user running the command lacks write permissions on /var/lib/etcd-new

D The data directory /var/lib/etcd-new already contains running etcd data

💡 Hint

Permission denied errors usually relate to filesystem access rights.

The error indicates the process cannot write to the specified directory due to insufficient permissions. Ensuring the user has write access to the data directory fixes this.

🔀 Workflow

advanced

Correct sequence to backup and restore etcd cluster

Arrange the steps in the correct order to safely backup and restore an etcd cluster:

A 2,1,3,4

B 1,2,3,4

C 1,3,2,4

D 2,3,1,4

💡 Hint

Backup first, then stop service before restore, then start service with restored data.

The correct workflow is to first create a snapshot, then stop the etcd service to avoid conflicts, restore the snapshot to a new data directory, and finally start the service using the restored data.

✅ Best Practice

expert

Best practice for automating etcd backups in production

Which approach is considered best practice for automating etcd backups in a production Kubernetes cluster?

A Schedule regular 'etcdctl snapshot save' commands with proper endpoint and authentication, storing backups securely off-node

B Backup etcd data directory files directly while etcd is running for speed

C Use etcdctl snapshot save without authentication flags to simplify automation

D Manually run snapshot commands only when cluster issues occur to avoid performance impact

💡 Hint

Consider security, reliability, and minimal disruption in production environments.

Automating backups with scheduled snapshot commands that include authentication and storing backups off-node ensures consistent, secure backups without manual intervention or risking data corruption.