Core & Module Updates via Composer won't pass pipeline to Pantheon

[Mark Jovanovich]

I am using Bitbucket to store the codebase for a pair of drupal sites that were designed by a 3rd party firm, which are hosted on Pantheon. I worked with the 3rd party to confirm that all proper SSH keys/machine tokens/etc. are in place and updates via a local clone w/ git/composer can run a pipeline through to Pantheon.

After my latest attempt to run drupal core/module updates, the pipeline now fails during the "Static Tests" & "Build PHP" stages, under "./.ci/test/static/run". It returns the following error: "

In CurlDownloader.php line 641:

195

196

The "https://codeload.github.com/drgullin/icheck/zip/1.0.2/tags/1.0.2" file

197

could not be downloaded (HTTP/2 404 )".

I checked the changes and there is no instance of "gullin" or "icheck" in composer.json, composer.lock, or the yaml file.
Moreover, after that I tried to just run the preceding pipeline which I knew worked, but now THAT pipeline is throwing the same errors. So now I don't have a working pipeline(?).
I also tried to "REVERT" the changes, both via my local git clone, but also via the Bitbucket Pull Requests menu. Neither worked. Via the menu, I will admit I don't know if I should select revert on the last PR that caused the issues, or select the preceding PR that worked and select revert on that one(?). At the time, it gave me a warning/error, saying that it would create a discrepancy because it was behind on files (?).
So now I don't know how to move forward. I know I would prefer to just kind of get back to where I was before the latest round of updates, so I can pull that through Pantheon again, while figuring out what about the updates cause the pipeline to break.
How can I address these issue? Any info is helpful and appreciated. Many Thanks.

[Ben]

Hi @Mark Jovanovich 

The first step I would recommend is to analyze your build setup and compare the SHA for the build image used for a working vs failing build. If these are different - it could be that a sudden image update has caused the problem, in which case - you can copy the full image name (with SHA) from the working build logs into your YAML and attempt to use this to execute a fresh build.

If this doesn't resolve the issue, we will need access to your build logs/YML config so we can analyze your environment and troubleshoot further.

To do so, we need a formal support ticket so our systems will allow us to grant the necessary access. Please raise a ticket in our support portal accordingly:

• https://support.atlassian.com/contact/

If you encounter difficulties doing so, please let me know your timezone so we can raise one on your behalf.

Cheers!

- Ben (Bitbucket Cloud Support)