Shibboleth/Jetty timeout

[Hrvoje Hodak]

Hello,

after testing our failover cluster for attributes, I've restarted our 
long running shibboleth-idp-5.1.3 via jetty and all of a sudden jetty 
didn't want to start up.

Jetty is not throwing any warnings or errors (on debug) except:

> State Parent Directory: drwxr-xr-x. 13 jetty jetty 159 Aug 19 13:52 /opt/jetty/jetty-base
> .State (now): STARTING oejs.Server at 1e178745{STARTING}[12.0.14,sto=5000]
> .State (now): STARTING oejs.Server at 1e178745{STARTING}[12.0.14,sto=5000]
> [...]
> .State (now): STARTING oejs.Server at 1e178745{STARTING}[12.0.14,sto=5000]
> .State (now): STARTING oejs.Server at 1e178745{STARTING}[12.0.14,sto=5000]
> Timeout 30 expired waiting for start state from /opt/jetty/jetty-base/jetty.state
>  timeout
> INFO: Server process is running
> FAILED Tue Aug 19 16:00:15 CEST 2025
> .Killed 18513

I've also tried increasing the timeout to 300 seconds, but unfortunately 
that didn't help.

Strangely, during that time, the idp seems to be running but no 
connection is possible. I can see Hikari starting up

 > INFO [com.zaxxer.hikari.HikariDataSource:122] - HikariPool-5 - Start 
completed.

but there are no relevant warnings or errors in the idp log.

Something seems to be causing the timeout and I'm at  a loss about what 
it could be. I thought the culprit could be our attribute-storage, but 
our other idp-node is working fine.

What am i missing?

Thanks in advance,
Hrvoje Hodak