Easier debugging ideas about container-common-scripts HOT 5 OPEN

How the keep-alive could work: A simple solution would be to add sleep to the cleanup script:

$ keep_time=$(echo "${ghprbCommentBody}" | grep -o '\[keep-machine[^]]*\]' | sed -e 's/\[keep-machine:\([^]]*\)\]/\1/')
[ -n "${keep_time}" ] && echo "Keeping the machine $IP alive for ${keep_time}" && sleep "${keep_time}"

Of course this can be enhanced by sending a github comment, tagging the person who added that comment etc...

from container-common-scripts.

Do you want to keep the VM allocated by resalloc alive, or the jenkins
machine? If the former, new resalloc feature will be shipped in v3.0;
you could put the machine into sandbox by --sandbox ... and anyone
(perhaps person doing the debugging) could re-take the machine with
the same --sandbox option (within pre-configured time-limit).

from container-common-scripts.

the docs

from container-common-scripts.

@praiskup wow, that helps, thanks. The original idea was very simple -- the job would wait and thus block the chain (there is a limit of 4 parallel jobs I think) -- so doing it through resalloc would be much better approach.

from container-common-scripts.

This would be really awesome. I am a volunteer for testing. :)

from container-common-scripts.