RE: Packaging 7.1.1 - Mailing list pgsql-hackers

From Rachit Siamwalla
Subject RE: Packaging 7.1.1
Date
Msg-id 9AC41B8C4781464695BB013F106FCA31D2BD29@nasdaq.ms.ensim.com
Whole thread Raw
In response to Packaging 7.1.1  (Bruce Momjian <pgman@candle.pha.pa.us>)
Responses RE: Packaging 7.1.1  (Rocco Altier <roccoa@routescape.com>)
List pgsql-hackers
Thanks a lot for your total and complete description of the process. (i
should have checked out the sprm first before asking).  I empathize with
what you said about packaging not being a simple task, i have been through
the agony.

About putting your stuff into the postgres tree, i believe it would be a
"good" thing other than bad to include it in pgsq. It can be put into the
contrib directory (because it isn't part of the "core" portable stuff). This
solution was done for the portable openssh cvs tree. not only redhat
packaging stuff was included, but the solaris pkg mechanism was also in
there (and i also believe there were some others). It usually isn't a lot of
files (ie. the spec file and maybe the initscript). Of course its up to the
gods of the pgsql tree what they want to do with it, so i'm just going to
raise this suggestion and shut up.

anyways, getting back to the what brought me to ask about this, can you add
the fixes to these two small problems in your initscripts?

1. `pidof` should be `pidof -s` (2 instances)
2. restart) should be stop; sleep x; start
ideally, stop should actually wait till postgres fully stops. The sleep is
just a temporary fix.

I have a more thorough email i sent earlier, i can resend it to you if you
want.

-rchit


pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: Re: New Linux xfs/reiser file systems
Next
From: Bruce Momjian
Date:
Subject: Re: Re: New Linux xfs/reiser file systems