From: | Alex Pilosov <alex(at)pilosoft(dot)com> |
---|---|
To: | Paul Tomblin <ptomblin(at)xcski(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: What the heck is happening here? |
Date: | 2001-06-11 22:15:16 |
Message-ID: | Pine.BSO.4.10.10106111813310.232-100000@spider.pilosoft.com |
Views: | Whole Thread | Raw Message | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Mon, 11 Jun 2001, Paul Tomblin wrote:
> Quoting Alex Pilosov (alex(at)pilosoft(dot)com):
> > In mod_perl case, make sure that your stuff does not try to establish
> > connections on startup, and only does that (using ApacheDBI to cache
> > connections) when the page is requested.
>
> Oh oh. I *am* doing that - I make a connection to grab some stuff right
> at start up, and keep that $conn for all the pages. Is there a good
> example of using ApacheDBI to cache connections using Pg.pm I can look at?
Don't do it. If you _must_ get some stuff on startup, close the conn after
you are done. See mod_perl references for examples on using of ApacheDBI
(mostly its just 'use ApacheDBI' vs 'use DBI') and no, you cannot use Pg
(the oldstyle, libpq-alike) interface with ApacheDBI, you must use
DBD::Pg.
> > Also, it is possible that every day some script tries to rotate logs and
> > SIGHUPs the httpd. mod_perl may do strange stuff on restart. Try manually
> > sending HUP to server and see what happens.
>
> Actually, sending HUP to all the httpd processes is the only thing that
> fixes it. For a while.
-alex
From | Date | Subject | |
---|---|---|---|
Next Message | Paul Tomblin | 2001-06-11 22:17:31 | Re: What the heck is happening here? |
Previous Message | Andrew W. Schmeder | 2001-06-11 22:12:58 | pg_dump problem... |