Replacing a Replication Gateway from a gateway pair
Using the Resiliency Platform console you can replace an existing Replication Gateway from a pair irrespective of whether the gateway is healthy or faulted.
To replace a gateway, you must go to the data center having the gateway and perform the replace gateway operation. In this case the peer gateway must be healthy.
If you want to replace both the gateways, then after replacing the first gateway wait till the operation is successfully completed before you replace the other gateway. If both the Replication Gateways are faulted, you cannot perform the replace Replication Gateway operation. You need to remove them and create a new pair.
Note:
1. Make sure you do not replace the Replication Gateway and its peer gateway from the Replication Gateway pair simultaneously. This may lead to replace Replication Gateway failure and you will have to delete the Replication Gateway, the gateway pair and its associated resiliency groups.
2. While replacing the Replication Gateway on the VMware environment, DRS automation level for the existing Replication Gateway and new Replication Gateway should be set to manual or should be disabled.
3. If you have replaced a Replication Gateway which is associated with a resiliency group which is protecting a physical workload, it is mandatory to perform resync operation once the replace gateway operation is successful.
Also, you can replace a gateway with another gateway only if it is configured to use the same mode of replication. For example, you cannot replace a gateway that is configured to use Direct mode of replication with the one that is configured to use Object Storage mode of replication.
When you replace a gateway, all the Veritas Replication Sets are moved to a new gateway and the Replication Gateway pairs are reconfigured. If the existing Replication Gateway is configured as a PXE boot server or as a DHCP server, then the new gateway is reconfigured as a PXE Boot server or a DHCP server.
Note that after replacing a gateway, irrespective of the replaced gateway being faulted or healthy, the replication resumes from the last successfully replicated update set.
Before you replace a replication gateway, if any of the resiliency groups that are associated with the gateway are in offline state, because the migrate or take over operation has failed, then you need to start the resiliency group before replacing the gateway. Ensure that you select the Refresh storage, network, compute and customizations check box in the start resiliency group wizard.
If the staging disk attached to the Replication Gateway is corrupted or removed, you need to shutdown the gateway and then perform the replace gateway operation.
After replacing a gateway, the Server Gateway attribute is inherited from the previous pair. If the Replication Gateway being replaced itself is the Server Gateway, then the attribute is updated to point to the new gateway. Ensure that the networking constraints between the two gateways are maintained with the new gateway.
Replace Replication Gateway operation is not supported in the following scenarios:
When recovering VMware virtual machines from an on-premises data center to vCloud Director.
When recovering virtual machines from vCloud Director to vCloud Director.
From version 3.5, you can configure the resiliency group for Continuous Data Protection (CDP) by selecting the Enable Continuous Data Protection checkbox while selecting the appropriate gateway pair.
To replace a Replication Gateway from a gateway pair
- Navigate to Settings (menu bar) > Infrastructure > Details View
You can also access this page from the Quick Actions menu.
- Click Data Mover > Resiliency Platform Data Mover tab.
- In the Replication Gateways table, right-click the Replication Gateway which you want to replace, and select Replace Gateway.
- From the list of gateways, select the replacement gateway, and click Next to initiate the replace operation.
The replace Replication Gateway operation may fail if there are any resiliency groups in configuration failed state, that are associated with the gateway. You then need to delete the Veritas Replication Sets, then delete the resiliency groups, and then retry the replace gateway operation.
While replacing a Replication Gateway having CDP storage attached, make sure that the datastore on which the CDP storage resides is accessible to new Replication Gateway.
More Information