Re: BUG #2236: extremely slow to get unescaped bytea data - Mailing list pgsql-bugs

From Kalador Tech Support
Subject Re: BUG #2236: extremely slow to get unescaped bytea data
Date
Msg-id 43EB7080.8060909@kalador.com
Whole thread Raw
In response to Re: BUG #2236: extremely slow to get unescaped bytea data  (Michael Fuhr <mike@fuhr.org>)
List pgsql-bugs
You guys are smart!

Apache was running against an old libpq.  I shutdown apache, updated
/etc/ld.so.conf with the postgres lib dir, ran ldconfig, restarted
apache, and the problem went away.

The old libpq was libpq.so.3.0 (pre-installed on machine).  The new one
is libpq.so.3.2 (installed with 8.0.1)

Sorry for the false alarm - thanks for the help.

Kai Ronan
Technical Support
Kalador Entertainment Inc.



Michael Fuhr wrote:

>On Thu, Feb 09, 2006 at 12:46:46PM -0300, Alvaro Herrera wrote:
>
>
>>I note in the PHP 4 sources that the PQunescapeBytea function seems to
>>have been copied there, "for the benefit of PostgreSQL 7.2 users".  It
>>says that it comes from 7.3 but I don't see any sscanf call.
>>
>>There is no PQunescapeBytea call in the whole source that I can see, so
>>my guess is that the libpq function is not called at all.  So this may
>>be a PHP bug rather than a Postgres bug.
>>
>>
>
>The OP claimed to be using PHP 5.1.2, which does have a call to
>PQunescapeBytea(), although it also has the old code you're seeing
>and a HAVE_PQUNESCAPEBYTEA macro that determines which to use.
>
>Interesting that the command line php and the Apache module behave
>differently.  I wonder if ldd would show the php executable and
>libphp5.so linked against different versions of libpq; that would
>add weight to Tom's suggestion that an old libpq might be responsible.
>
>
>

pgsql-bugs by date:

Previous
From: Michael Fuhr
Date:
Subject: Re: BUG #2236: extremely slow to get unescaped bytea data
Next
From: Kris Jurka
Date:
Subject: Re: BUG #2250: JSTL parameterized queries inserting numeric