Applies toOCI Block Volume (boot and block volumes, volume groups), cross-region replication and backup copy
Options
| Method | RPO | RTO | Cost |
|---|---|---|---|
| Cross-region replication | Typically under 30 minutes (asynchronous, continuous) | Minutes (activate replica, launch instance) | Replica storage in DR region |
| Cross-region backup copy | Backup frequency (e.g. daily) | Restore time (size dependent) | Backup storage |
| Application replication (Data Guard, DB-native) | Seconds | Minutes | Running standby |
1. Replicate a volume group (consistent set)
Put the boot volume and data volumes of an instance (or of several related instances) into a volume group, then enable replication:
oci bv volume-group create --compartment-id <cmp> --availability-domain "Abcd:AP-MUMBAI-1-AD-1" \
--display-name vg-app01 --source-details '{"type":"volumeIds","volumeIds":["<boot-vol-ocid>","<data-vol-ocid>"]}'
oci bv volume-group update --volume-group-id <vg-ocid> \
--volume-group-replicas '[{"availabilityDomain":"Efgh:AP-HYDERABAD-1-AD-1","displayName":"vg-app01-replica"}]'
Volume group replication keeps the volumes crash-consistent with each other, important for apps that spread data and logs over multiple volumes.
2. Check replication
oci bv volume-group-replica list --compartment-id <cmp> --availability-domain "Efgh:AP-HYDERABAD-1-AD-1" \
--query "data[].{name:\"display-name\",state:\"lifecycle-state\",synced:\"time-last-synced\"}" --output table
Alert if time-last-synced is older than your RPO.
3. Fail over
- In the DR region, activate the replica (creates a usable volume group from the latest synced point).
- Launch an instance from the activated boot volume and attach the data volumes (or let FSDR do it).
- Fix IPs/DNS, start the application, validate.
4. Backup copies (cheaper DR)
Backup policies can include a cross-region copy schedule. Restoring creates new volumes from the copied backup in the DR region. Good for Tier 2 systems with RTO in hours.
oci bv boot-volume-backup copy --boot-volume-backup-id <backup-ocid> --destination-region ap-hyderabad-1
5. Things to plan
- Encryption keys: replicated volumes in DR use a key you specify in the DR region's Vault. Make sure the key exists and policies allow it.
- Performance: set VPUs on the activated volumes to match production.
- Replication is storage-level: databases on these volumes recover like after a crash; prefer database-native replication for critical databases.
Tip
Combine volume replication with Full Stack DR so activation, instance launch and LB updates run as one tested plan instead of a manual checklist at 3 AM.