by Luc Saffre · in Developers · assigned to hannes

Résumé

Concepts sous-jacents

Commentaires

• 2 ans, 8 mois ago by hannes about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : I have some other idea: In gitlab, Settings -> CI/CD -> General pipelines there is Timeout Jobs fail if they run longer than the timeout time. Input value is in seconds by default. Human readable input is also accepted, for example 1 hour. I changed this to 4h. Lets see, how it behaves with this ne...
• 2 ans, 8 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : The issue is again there ([url gitlab.com/lino-framework/presto/-/jobs/6039924328 example]). @hannes can you write a cron job that does this automatically? Or do you have other ideas?
• 2 ans, 10 mois ago by hannes about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : du -ksh /var/lib/docker/ 7,2G /var/lib/docker/ # service docker stop # cd /var/lib/docker # rm -rf * # service docker start # du -ksh /var/lib/docker/ 260K /var/lib/docker/
  
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : I don't t believe that we should force things. But what means that a container is up? Can't we prune a container as soon as GitLab has read the results?
• 2 ans, 10 mois ago by hannes about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : docker system prune --volumes can't remove volumes, if these are in use, e.g. container is up. Maybe option --force helps here.
• 2 ans, 10 mois ago by hannes about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : tests/test_demo.py::Main::test_noi1e FAILED [ 8%] happened again
• 2 ans, 10 mois ago by hannes about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : /dev/sda1    20470152 19459236  141328 100% / Diskspace of kuu is now almost full.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : @hannes please research why "docker system prune --volumes" removed only 4 images with a total of 2 GB and why docker is still using some 9 GB.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : See also #5295, both tickets are probably related.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : The timeout in /etc/gitlab-runner/config.toml is 18000 seconds, which means 5 hours. That's much more than needed.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : "docker system df" said that there were 11 images and 52 volumes taking 9GB of disk, 100% of which "reclaimable". But after "docker system prune --volumes" it removed only 4 images with a total of 2 GB. It is still using some 9 GB.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : But where can I set the timeout limits?
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : Happened when building [url gitlab.com/lino-framework/welfare/-/jobs/5791219725 welfare on gitlab]. Maybe it just took really too much time.
• 2 ans, 10 mois ago by Luc Saffre about #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again) : Created new Ticket #5337 (There has been a timeout failure or the job got stuck. Check your timeout limits or try again).