Docs menu
MENU

CloudBase Coriolis Disaster Recovery Guide

CloudBase Coriolis offers Disaster recovery as a Service (DRaaS) through Coriolis Replicas.

coriolis dr

coriolis features

When using Virtuozzo Infrastructure as the source environment for disaster recovery, Changed Block Tracking (CBT) is not supported. For virtual machines with large disks (for example, 1 TB or larger), this may result in increased storage load during replication.

Support for CBT with Virtuozzo Storage is planned for a future Coriolis release.​

Replica setup

Replica executions represent scheduled background sync of a running workload’s storage from a source cloud directly to a destination cloud.​

Replica deployments are the ability to recreate the source VM on the destination cloud using the synced storage, regardless if the source platform is still active or not.​

1. Set up replication for the desired VM, as described in the CloudBase Coriolis Migration Guide.

2. Configure the replication schedule in the replica settings. Recovery point objective (RPO) depends on how often you perform replication and how long it takes for a replica to execute, so you can set RPO to the required minimum. For example:

  • If you need to execute replicas once an hour, it is enough to have just a single schedule:

coriolis replica setup1

  • If you need to execute replicas every 15 minutes (4 times per hour), then you need to configure 4 schedules:​

coriolis replica setup2

Schedules can be created during replica configuration or after an initial replica execution. A schedule performs incremental syncs that take less time since only the modified blocks are synced.​​

To better understand Coriolis RPO/RTO, refer to Coriolis Recovery Point Objective & Recovery Time Objective.

Failover

When the source Site A fails, you need to create migration from a VM replica to create a VM on the destination Site B. At this point, Coriolis will create a VM from the replicated volume on the destination site:

coriolis failover

Failback

When Site A is recovered, you can create failback migration to move the failover VM back from Site B to Site A. To do this, create one-step migration, as described in CloudBase Coriolis Migration Guide | One-step migration:

1. Select Site B (Virtuozzo Infrastructure) as the source endpoint.

2. Select the recovered VM.

3. Select Site A (Virtuozzo Infrastructure) as the target endpoint:

3.1. Specify the VM flavor, network, and disk storage policy.

3.2. Switch the Shutdown Instances option to ON to stop the source VM.

4. Execute failback migration.