Gitlab Runner Kubernetes Executor Cache Not Working

Hello everyone,

I’m managing a Gitlab Core Edition on one server I’m renting.

We are currently running 13.2.4-ce Gitlab version.

I try to migrate our docker runner to Kubernetes with the Kubernetes Executor on Gitlab CI Runner.

For a simple job, everything work well, but I can’t have the cache working.
(The caching system was working well on docker previously)

When the job tries to save the cache, I receive this error:

Saving cache
Creating cache ee42024c36f22d366b965fb06d46bf07a4a9df6d…
ERROR: Job failed: command terminated with exit code 137

I don’t see any logs anywhere (pods / containers / machine /var/log/syslog…)

How can I debug it to diagnose the problem?

I found the error :slight_smile:

I was setting up some Mem and CPU limits and requests for the Gitlab Runner, unfortunately, the Cache step needed more CPU & Mem than the limit and the container was killed by Kubernetes without any log.

If anyone has the same problem… Check your requests & limits :slight_smile: