Hi Adrian, Adrian Bridgett [2006-03-21 11:35 +0000]: > /etc/init.d/postgresql-8.1 status always returns 0 (it just backs this > call off to /usr/share/postgresql-common/init.d-functions hence the > bug on postgresql-common). > > This causes issues with HA since it requires LSB compliant scripts - > in particular status should return "3" if postgresql is down.
Thanks for your report. It's not trivial, since you can have arbitrarily many clusters which can be running, or down (even deliberately) independently of each other. > I've attached a short patch to do this, however it's not as clean as > it should be It's not easy to define a semantics for that init script which makes sense. However, in any case I can add a 'status' command to pg_ctlcluster which returns LSB conformant exit codes, so that there is a tool to check a particular cluster. This could then be integrated into the init script somehow. > since there is a global "exit 0" in the > /etc/init.d/postgresql-8.1 script (why?!). That's common practice for init scripts, since they usually abort immediately on errors (due to set -e). I just saw that the postgresql-X.Y scripts forgot this, thanks for making me aware of that. :) Martin -- Martin Pitt http://www.piware.de Ubuntu Developer http://www.ubuntu.com Debian Developer http://www.debian.org In a world without walls and fences, who needs Windows and Gates?
signature.asc
Description: Digital signature
