Ticket #934 (closed defect: wontfix)
bcfg2 hangs waiting for defunct init script
Reported by: | https://www.google.com/accounts/o8/id?id=AItOawnSjgovXZr-_V3vGkvMSR0pc5LDykRc1Nc | Owned by: | desai |
---|---|---|---|
Priority: | minor | Milestone: | Bcfg2 1.2.0 Release |
Component: | bcfg2-client | Version: | 1.0 |
Keywords: | Cc: |
Description
We have an init script for Osiris (http://osiris.shmoo.com/) that, when run by bcfg2, goes defunct, which hangs bcfg2 until we do a Ctrl-C. I have attached the init script we are using, since it likely has some local modifications.
We are using bcfg2 1.0.1 and Osiris 4.2.3.
Attachments
Change History
Changed 13 years ago by https://www.google.com/accounts/o8/id?id=AItOawnSjgovXZr-_V3vGkvMSR0pc5LDykRc1Nc
comment:1 Changed 13 years ago by https://www.google.com/accounts/o8/id?id=AItOawnSjgovXZr-_V3vGkvMSR0pc5LDykRc1Nc
It appears that this is _not_ a bug with bcfg2, but something with the way osirisd handles stdout and stderr. I've replicated the issue with Python and Perl, completely separate from bcfg2.
As rare as it is, it'd be nice if bcfg2 could handle this sort of thing, but I'm betting there's not much Python can do if an external program misbehaves like this -- at least, not trivially. This is safe to close as notabug. (I would if I could figure out how....)
osiris init script