How to solve a stuck Gitlab CI pipeline?

Question

We've been using Gitlab CI for some months, and in the last 1 week, we've been using the specific runner installed on a VPS. Currently, we are using "shell" as the executor.

Today our pipeline got stuck out of sudden, when we looked into the server free RAM, it's only 48MB out of 996 MB, FYI, we're using CentOS 6.

We've been struggling to get the answers, but we're stuck at the moment, and would like to know :

• What's causing the pipeline from getting stuck?
• is it true because of low free RAM?
• Should we use another executor, perhaps SSH or even docker?
• What is the best practices to deal with this kind of problem?

We would appreciate any kind of help or directions.

Answer 1

In my case the pipeline was stuck because the only available runner had the option "Can run untagged jobs" set to "No" and the job was really untagged. One can fix this issue by changing the "Can run untagged jobs" option or by adding a tag to the appropriate section of the ".gitlab-ci.yml" file in the repository. In my case it was section default:tags:.

(It seems that your case is much more complicated. However I've came across this issue twice a month, and I've forgotten the decision at the second time. Thus I've came to this page which looks appropriate to save the decision. Hope the answer will help someone else.)

Answer 2

In my case, the pipeline was stuck because of two things:

1. The tags specified in the .gitlab-ci.yml do not match those in the runner configuration.

2. If you specify the simulator in the build command, ensure that you write the right version of the simulator.

Once I did these changes, everything worked well!

Good luck.