From Jenkins, hostnames are used, not static IPs, both on CD and CI.

Decreasing the priority, as I've deployed PHP agents with D3602 fix to docker-002.

Apparently some Jenkins agents have static ip. But that ip can be taken by dynamic assignment of previously launched containers.

Alkane will now PULL files from Jenkins, instead of Jenkins pushing to the relevant server.

https://devcentral.nasqueron.org/feed/7214929457763397007/

Next: publish this somewhere

https://github.com/nasqueron/spectacle/commit/82eca117830194264c0976d1713caab25803dc09 works fine with dart-sass and grunt-sass

https://sass-lang.com/dart-sass is compatible with node-sass API (read about fibers).

A good candidate for this one could be rDS

First plan of action could be to ensure we've effectively and explicitly in the job configuration moved to Next Generation Warnings plugin.

The other one is reported upstream - https://github.com/theseer/phpdox/issues/424

The xdebug one is fixed by this commit, that's why it doesn't occur on WindRiver:

Issue has been fixed in 2022, see https://github.com/jenkinsci/publish-over-ssh-plugin/pull/34

GitHub repository requires authentication, https://verdaccio.org/docs/cli-registry gives solution to prepare a private registry.

As a temporary solution to fix Spectacle, I've upgraded the dependencies [1] and published a package to GitHub packages repository [2].

I've rebuilt and push to Docker Hub a fresh set of jenkins-agent-* images.

DevCentral update: Updated tasks and passphrases to use "agent" word.

Let's try in the container by the way:

About phantomjs and silly_barden:

All seem in order, including the xsl extension.

Now fully operational since end August.

3. refactoring done.
4. filtering partially done: only failure are reported, Jenkins notifies, we consume.
5. to automate without human assertion system status isn't currently considered as valuable

Step 4 done for B to D, but currently every failure will be reported. And we don't exploit in the artefact log what's failing.