Is this a new installation, or a working one that has suddenly starting acting strange?
How much memory do you have on each of the servers?
What does Tabadmin status -v say?
I think that you are going to have to wait for Tableau Support to look through the logs and come back to you
All the best
In addition to what Glen asked.
- What release of TS are you running?
- Are the nodes running on VM, physical, or in the cloud?
- Have a look at event viewer, perfmon, and the tabadmin logs on each of the nodes to figure out what is going on.
Just to circle back on what we found for this problem (for posterity's sake :-)
Machines: VM with WIndows Server 2012 R2
Tableau Server: Version 9.3
It turned that during the configuration of the worker nodes, the sever configuration manager was not able to create all the inbound rules in Windows Firewall that it needed to allow the connections between the primary and the workers. These rules are for executable files and not for ports. The strange thing was that it created some of them, but not all of them. And each environment had different rules missing. There was nothing in our domain policies or VM configuration that should have prevented it from allowing the creation of those rules, but for whatever reason, they were randomly omitted. I still don't have an explanation for it, but I was able to correct it.
Simply copying the missing inbound firewall rules from a node that had them to the nodes that didn't. That stabilized the environment and fixed the problem. It's working great now.
Hope that helps someone in the future.