jenkins-plugin-cli reports War not found, installing all plugins: /usr/share/java/jenkins.war

[dominykas]

Jenkins and plugins versions report

Official container version: 2.439-alpine

What Operating System are you using (both controller, and any agents involved in the problem)?

Alpine

Reproduction steps

1. Run jenkins-plugin-cli -l

Expected Results

Lists the installed plugins, without any warnings.

Actual Results

Prints War not found, installing all plugins: /usr/share/java/jenkins.war at the start.

Anything else?

This is because the jenkins-plugin-cli changed the default Jenkins WAR path in jenkinsci/plugin-installation-manager-tool#622 (released as 2.12.14, first appeared in 2.436 [2.435 seems to still be OK, despite Github claiming this was released in 2.426.2] via #1794)

Are you interested in contributing a fix?

I'm not sure what the fix should be, but I could give it a stab.

[MarkEWaite]

I think the fix should be made in the Jenkins plugin installation manager tool. It should check both the old and the new locations rather than only checking the new location

[dduportal]

I think the fix should be made in the Jenkins plugin installation manager tool. It should check both the old and the new locations rather than only checking the new location

I vote for rollbacking the change in the CLI and republish it under a 3.x version as it looks like a breaking change

[hervelemeur]

An additional test checking the absence of "War not found, installing all plugins: /usr/share/java/jenkins.war" would be nice too.

[timja]

An additional test checking the absence of "War not found, installing all plugins: /usr/share/java/jenkins.war" would be nice too.

checking for absence of isn't that useful, it should verify that whatever is broken works correctly here though. I'm surprised it got through the tests here 😢.

[ninjonxb]

Since this just caused us significant problems and a broken Jenkins state even after downgrade (jenkinsci/plugin-installation-manager-tool#641 )

I would like to propose that this check for the previous default location as well before throwing an error? I would not think that both would exist on a system and if so just default to using the new default?

Or at the very least a comment to make sure this works with the official docker image since this would be a breaking change that is a bit hard to track down.