One or more nodes are not accessible from the vRealize Automation instance or
There is some miscommunication between the Management Agent and its data within the vRealize Automation database.
Resolution
To resolve this issue, follow these steps:
Log on to all machines where the certificate failed to install and manually upgrade the Management Agent and try again.
If the failing node is not reachable because it has been removed previously:
Snapshot/ backup your vRealize Automation IaaS servers.
Try manually installing the vRealize Automation 7.1 Management Agent on the failing server and try again.
If the unreachable nodes are Windows servers, then vRealize Automation Management Agent service should be running on them.
Check the network connectivity (Ping) between the node(s) and the appliance you are upgrading from.
DNS should be configured correctly (nslookup) between the node(s) and the appliance you are upgrading from.
Review the Management Agent logs for any errors that may provide more detail on why the failure is occurring.
If the node is invalid or is removed previously from the environment, remove the node entry from the primary vRealize Automation appliance:
Note: If you are running on vRealize Automation 7.0.1 or later, navigate to vRA Settings > Cluster tab in VAMI page and click Delete in the Nodes table to remove the invalid nodes.
Back up your vRealize Automation appliances and Postgres Databases.
SSH into the primary vRealize Automation appliance and run these commands in the order specified:
cd /opt/vmware/vpostgres/current/bin
su postgres
./psql vcac
After logging in to the postgres database, run these commands in the order specified:
\c vcac
\x
select * from cluster_nodes;
Locate the node_id(s) for the nodes you want to remove and run this command:
delete from cluster_nodes where node_id = '<ID from previous step>';