Tableau Server Multi-node Cluster with Site-Specific SAML and Additional Node with Primary Repository Goes Down Users Cannot Sign In After Node Recovery

Published: 03 Oct 2022
Last Modified Date: 24 Oct 2022


In a multi-node Tableau Server with site-specific SAML configured, when an additional node that contains the primary repository goes down, users cannot sign in even after the node has recovered.  After the starting up the node which was down, site-specific SAML sign-in keeps failing with error:

This page isn't working HTTP ERROR 500


  • Tableau Server 2022.1
  • Multi-node HA cluster
  • Site-Specific SAML


To work around this issue, please end the samlservice.exe once from task manager on the remaining nodes if they are running samlservice.exe.

User-added image

Then confirm that the samlservice.exe automatically starts up with another PID as demonstrated in the screenshot below:.
User-added image

If the failure node can be recovered and can be started up, please restart Tableau Server processes by tsm restart after the failure node has recovered.


This issue is currently under investigation with ID 1439848.

Additional Information

This issue does not occur in Tableau Server version 2021.4.
Did this article resolve the issue?