Confluence node had problems starting up after upgradation

[Kimmi]

While upgrading the Confluence data center from 8.5.20 to 9.2.6, 2 out of 3 nodes started without any problems, but Node 3 had some problems starting.

2025-08-28 19:43:05,622 ERROR [http-nio2-8443-exec-4] [atlassian.confluence.util.AbstractBootstrapHotSwappingFilter] doFilter Could not get swap target filter
-- url: /confluence/rest/capabilities | traceId: 00f835aa2947f2d3
2025-08-28 19:43:05,624 ERROR [http-nio2-8443-exec-4] [[Standalone].[localhost].[/confluence].[noop]] log Servlet.service() for servlet [noop] in context with path [/confluence] threw exception
2025-08-28 19:43:05,625 ERROR [http-nio2-8443-exec-4] [[Standalone].[localhost].[/confluence].[system-error]] log Servlet.service() for servlet [system-error] threw exception

[Rajat Pratap Singh]

Hi @Kimmi 

Have you copied the same Confluence home directory as you copied for the 2 working nodes?

If yes, please try the following:
1. Clear the plugin cache:
Stop this node 3
Delete the plugin cache on this node:
<confluence-home>/bundled-plugins
<confluence-home>/plugins-cache
<confluence-home>/plugins-osgi-cache
<confluence-home>/plugins-temp

Try starting the node now and check if it works.

2. Check if there is problem for the third node to join the cluster
Compare confluence.cfg.xml if using TCP/IP discovery across all nodes.
Make sure the node 3 can communicate with the other two nodes on the cluster port (default 5701).

3. Verify the database connectivity from node 3
Even though all nodes point to the same DB, sometimes the third node’s firewall could be blocking the connection.

4. Check java version compared to other nodes.

5. Check if node 3 disk is full.

Hope this helps!

Regards
Rajat

[Kimmi]

Hello @Rajat Pratap Singh 

Thank you for replying.

Mentioned point 1 and 4 has been checked already during upgrade.
Clearing cache did not helped. And java version is same for all nodes.

I will check other mentioned points.

Thank you!