0

On a VPS (that is not heavily used) with 4 shared vCPU's and 8GB RAM I'm running the following services:

  • Traefik for reverse proxy
  • Portainer for docker management
  • Gitlab, used by 5 people
  • Grafana with Influxdb for monitoring
  • Couple of Piwigo instances that serve a couple requests per day
  • WikiJs for 4 users

It does sound a lot but and InfluxDb receives the most data from remote sites.

Around every 4 a 5 days, the following error shows in the console of the VPS provider at the login screen:

Ubuntu 22.04.3 LTS morty tty1

ubuntu login: [500982.969952] Out of memory: Killed process 3155 (chromium) total-vm:1186888752kB, anon-rss:16412kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:676kB oom_score_adj:300

(500993.770320] Out of memory: Killed process 3156 (chromium) total-vm:1186886672kB, anon-rss:16416kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:660kB oom_score_adj:300

(501015.761075] Out of memory: Killed process 3154 (chromium) total-vm:1186886672kB, anon-rss:16412kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:664kB oom_score_adj:300

[501031.089045] Out of memory: Killed process 3188 (chromium) total-vm:33966888kB, anon-rss:15736kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:488kB oom_score_adj:200

(501048.747291] Out of memory: Killed process 3129 (chromium) total-vm:33967084kB, anon-rss:15688kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:488kB oom_score_adj:200

(501198.429539] Out of memory: Killed process 3130 (chromium) total-vm:33966888kB, anon-rss:15672kB, file-rss:0kB, shmem-rss:0kB, UID:1001 pgtables:488kB oom_score_adj:200

(502092.118577] Out of memory: Killed process 249356 (ruby) total-vm:1782896kB,anon-rss:1044976kB, file-rss:0kB, shmem-rss:0kB, UID:998 pgtables:3120kB oom_score_adj:0

(502093.377218] systemd[1]: snapd.service: Watchdog timeout (limit Smin)!

To me it looks like the error is caused by Gitlab, since it's mentioning Ruby. I don't know why it's mentioning Chromium.

I've never had this problem before, but it happened after I upgraded Gitlab to the latest versions (it was a couple of versions behind).

When this happens, the VPS becomes unreachable. Can't login anymore with SSH. When I login via the VPS provider there is also no network going outside. The only solution is to reboot the VPS.

Since it's mentioning memory error, will an upgrade to 16GB likely solve the issue? Or are there other solutions I can try.

According to Gitlab site:

Up to 20 Requests per Second (RPS) or 1000 users - 8 GB (Minimum), 16 GB (Recommended).

Since there are only 5 users, this seems unnecessary. Gitlab only hosts repositories, it's not used for any CI/CD. Maybe I could migrate to a different lighter Git server but I haven't looked into that yet (any recommendations?)

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.