Snapshots

A snapshot is a point-in-time, crash-consistent copy of a volume. Two common uses:

1. Backup — take snapshots on a schedule so a corrupted filesystem or accidental rm -rf is recoverable.
2. Golden images — install software once, snapshot, then boot every new VM with --root_volume_source_snapshot_id.

Snapshots are stored independently of the source volume. Deleting the volume does not delete its snapshots.

Create#

exc compute snapshot create --volume_id 7

The snapshot is taken without detaching the volume. It captures whatever is on disk at the moment the API call lands — application-level consistency (databases, anything with caches in RAM) is your responsibility. For databases, flush / fsync / xfs_freeze before snapshotting if you can’t tolerate replay on restore.

Use as a boot source#

exc compute create \
  --name web-2 \
  --image_id 1 \
  --instance_type m1a.large \
  --subnet_id 1 \
  --ssh_pubkey my-key \
  --root_volume_source_snapshot_id 99 \
  --wait

The new VM boots from a fresh volume cloned from snapshot 99. This is the cleanest way to roll out an immutable golden image.

Lifecycle#

exc compute snapshot list
exc compute snapshot delete --snapshot_id 99

Console#

1. Open console.excloud.dev/console/snapshots.
2. Review snapshot state, source volume, size, and creation time.
3. Use the snapshot row actions for available lifecycle operations.

Scheduling#

There’s no native scheduler today; pick one of:

• A cron job on a bastion that calls exc compute snapshot create.
• The Terraform excloud_snapshot resource, applied on a schedule by your CI.
• An Excloud Scheduler entry (when generally available).

Required permissions#

See the Permissions Reference.