Fix AccessLoggingIT when test is executed with long working directory path

[michalvavrik]

Summary

When AccessLoggingIT is executed in native mode in FIPS-enabled environment, working directory path in Jenkins can be long, which according to quarkusio/quarkus#44346 means that maximum log file size can get bigger than required size. Therefore I am applying workaround - making sure that path to log directory is short.

This test was added recently #2119 and the TD was backported, therefore this fix needs to be backported to 3.15 as well.

Update: now that I thought about it more I have doubts, see #2167 (comment).

Please select the relevant options.

• Bug fix (non-breaking change which fixes an issue)
• Dependency update
• Refactoring
• Backport
• New scenario (non-breaking change which adds functionality)
• This change requires a documentation update
• This change requires execution against OCP (use run tests phrase in comment)

Checklist:

• Methods and classes used in PR scenarios are meaningful
• Commits are well encapsulated and follow the best practices

[jcarranzan]

It looks good to me, thank you.
Are you run also checks on Openshift?

[fedinskiy]

@jcarranzan there are no OCP tests in this module: https://github.com/quarkus-qe/quarkus-test-suite/tree/main/logging/jboss/src/test/java/io/quarkus/ts/logging/jboss

[michalvavrik]

This PR was based on this quarkusio/quarkus#44346 (comment) suggestion. Now that I think about it I am not quite sure it makes sense because IMHO the issue is that this part:

 {"timestamp":"2024-11-05T15:57:06.816135Z","sequence":517,"loggerClassName":"org.jboss.logging.Logger","loggerName":"io.quarkus","level":"INFO","message":"Profile prod activated. ","threadName":"main","threadId":1,"mdc":{},"ndc":"","hostName":"rhel8-fips-xlarge-263","processName":"/home/hudson/hudson_workspace/workspace/rhbq-3.15-baremetal-ts-native/420ae479/quarkus-test-suite/logging/jboss/target/logging-jboss-1.0.0-SNAPSHOT-runner","processId":27639}

is bigger when it is native (because path to JDK is shorter than to native executable). Also I wonder why is it longer in FIPS, only diffrence I can see is that rhel8-fips-xlarge-263 contains extra fips :-D. Nevertheless, let's hope that expert is right and dummy (me) is wrong. I'll backport it and rerun.

TL;DR; if the route cause is same as the reproducer that I documented in quarkusio/quarkus#44346, then I don't know how this fixes it. It can only be tested in Jenkins so the easiest way is to simply run it. Will do it later today.

[michalvavrik]

I have tried it and I it fails. I think that suggestion was wrong, I'll comment on the issue. As far as fix goes, only workaround would be to place native executables to /tmp/something. I suppose there is easier way and it is to just avoid "processName" in there completely. I'll experiment with it and ask for new review.