There are multiple reasons when failover/failback replication is needed. For example:
This is a best method to perform radical Core and repository changes without affecting production environment and stopping backups for long time.
To perform failback/failover operation, the following prerequisites should be met:
The process consist of these phases:
1. Replicating all data to the target Core
Please perform the following configuration steps:
- setup replication link for all agents from the source Core to the target
- pause replication immediately
- opening both Cores' console, make sure that retention policies are matching. Both global and per-agent retention policies should match. If retention policies do not match, then Nightly Jobs on the target Core will rebuild RP chains accordingly and number of RPs for agents will not be the same.
To access global Retention Policy: Core console, More (...), Retention Policy
To access per-agent Retention Policy: Core console, click on agent name, click More dropdown on the right pane, Retention Policy
- once Retention Policies are matching, unpause the replication
Data will start replicating between Cores. Once all data is replicated, please verify that on both Cores by comparing number of RPs for each agent. If number of RPs is matching on both, then go the the step 2
2. Taking protection ownership from the target Core for all machines
Please make sure that source Core is no longer actively replicating any data at the moment, then pause protection for all agents on it. Also please cancel any active transfer jobs running.
- click on the source Core, Replication icon and delete outgoing replication link
- on the target Core, Replication icon, delete incoming replication link. Note: do NOT select option "with recovery points"
- on the target Core, protect all agents one by one. Agentless - Protect Multiple machines using VMware/HyperV/etc
3. Performing necessary steps on the source Core (installing new one, deprecating, rebuilding repository, etc)
4. Replicating all data back, from the target Core to the source Core
Now the process is essentially reversed as shown in the step 1, just Cores will replicate data back, from the target/temp Core back to the source Core, whether it's the same Core machine or newly built one.
- setup replication link from the target Core to the source Core
- pause replication
- if source Core was rebuilt from grounds up, please make sure to reapply same retention policy globally and per-agent
- unpause the replication and let the data flow back
5. Taking protection ownership back from the source Core
Once all data has been fully replicated back, please do:
- pause the protection on the target Core for all machines
- delete replication links on both Cores. Note: do NOT select option "with recovery points"
- on the source Core, protect all agents one by one. Agentless - Protect Multiple machines using VMware/HyperV/etc
At this moment, all data is stored on the source Core and agents are protected again. If target Core will continue to serve as a target, then replication link need to be setup from the source Core to target Core again.
Important: if encryption keys were used, please convert encryption keys from replicated to universal after each replication phase.
© ALL RIGHTS RESERVED. Feedback 利用規約 プライバシー Cookie Preference Center