Omnibus boot memory metrics job is failing
A while ago we introduced a step to the omnibus nightly builds that takes memory measurements of all booted components (omnibus-gitlab#3858 (closed)). As reported via Slack, this job has started to fail.
Hey Memory group, In omnibus-gitlab!3428 (merged), we added a job in omnibus-gitlab pipeline which spins up an instance using the GitLab docker image and get memory usage information from it. This job has been failing recently. Before I went in and chacked what broke and how to fix, I wanted to ask if Memory had any plans for this job at all which makes working on it worth it. If we don't have any plans to use the output from this job in the near future, I recommend we temporarily comment out this job (or make it manual) so as to save both money and developer time. WDYT?
https://gitlab.slack.com/archives/CGN8BUCKC/p1587274679309800
Error:
Gitlab services failed to start within 360 seconds
https://dev.gitlab.org/gitlab/omnibus-gitlab/-/jobs/7485180
Note: The job in nightly pipelines has been made manual via omnibus-gitlab!4130 (merged). Once a fix is in, we should revert this to be automatic