I was wondering if you were still able to replicate this problem in lucid? The current behaviour looks like it should result in the runsv processes remaining but the runsvdir process being killed.
I've added a debdiff against the runit source package in maverick here:
It should fix the issues with runsv being killed, but I can't explain why runsvdir would still be active. Killing runsvdir is handled by upstart itself.
Hi there,
I was wondering if you were still able to replicate this problem in lucid? The current behaviour looks like it should result in the runsv processes remaining but the runsvdir process being killed.
I've added a debdiff against the runit source package in maverick here:
https:/ /bugs.launchpad .net/ubuntu/ +source/ runit/+ bug/245728
It should fix the issues with runsv being killed, but I can't explain why runsvdir would still be active. Killing runsvdir is handled by upstart itself.