Cloud Director 10.2 when deployed in an HA configuration utilises at minimum a 3 node cluster composed of 1 x Primary Applaince plus 2 x Standby Appliances.

In the default deployment of 10.2 these nodes are deployed with a Manual failover of the Embedded Postgres Database. You can manually initiate a failover of the Database between nodes but for most enterprise customers you would want to set this to automatic so that failover happens without intervention.

In order to do set this in this version you must use the Cloud Director API to set the failover mode to Automatic.

Process

In this case I am using Insomnia as my API client.

Set HTTP Request to POST and input the Request URL as follows:

http://vcd01.terataki.local:5480/api/1.0.0/nodes/failover/automatic

Ensure that under the Authentication you select Basic Auth and input valid credentials. Here I am using Root for the cell.

Under the header section ensure that you have Accept|application/json as an input. No further headers are required.

The API call can be run against any node in the VCD cluster. As you can see I am running it against VCD01 which happens to be my primary node. Click Send to submit the request

This should result in a 202 Accepted Message. If you have any failures check that you have set the username/password and header correctly.

Checking in the VAMI of any of the nodes should now show that the failover mode for the cluster is now set to automatic.

Limitations

One thing to note here is that if you add additional nodes to the VCD cluster in future they will not inherit the setting of automatic. You must rerun the above command to update the fail over for all nodes.

Documentation

Official Documentation regarding the setting of the failover mode and the limitation mentioned above can be found on the VMware Cloud Director 10.2 Documentation website.