Re: BUG #4096: PG 8.3.1. confused about remaining disk space

----- Original Message -----
From: "Zdenek Kotala" <Zdenek(dot)Kotala(at)Sun(dot)COM>
To: <j6m(at)adm(dot)estp(dot)fr>
Cc: <pgsql-bugs(at)postgresql(dot)org>
Sent: Monday, April 07, 2008 12:51 PM
Subject: Re: [BUGS] BUG #4096: PG 8.3.1. confused about remaining disk space

> j6m(at)adm(dot)estp(dot)fr napsal(a):
>
>> I ran today a legacy application with PG 8.3.1 on a test machine.
>>
>> The DB is some ~30 MB. Default tables do have OIDs and BLCKSZ is set
>> system-wide to 16384 (because of another DB).
>
> Can you tried your application with standard BLKSZ? Or Can you send
> testcase?
>
>>
>> A lot of messages like these do appear wich tend to indicate PG 8.3.1 is
>> confused. This set of SQL scripts used to run from PG 7.0.3 to PG 8.2.6
>> with
>> no problem.
>
> What was BLCKSZ on these servers?

---8<-- SNIP

>
> User quota or super user space reservation could take affect. Could you
> create file as a postgres user on pgsql filesystem?
>

BLCKSZ was 16384 too. Actually, it ran with no problem on PG 8.2.6 with the
very same Linux distro (openSuSE 10.3) on the very same machine. So, it
does not seem the system settings are faulty (permission, user quotas, etc.
etc.).

Last week, I was running POG 8.2.6 on it. When 8.3.1 went out, I decided to
give it a try. I did pg_dumpall, recompiled PG 8.1.3 with the options that
were given by pg_config --configure and restored the dumpall file.

Needless to say, all regression tests were OK and I also checked the
libpq.so symbolic links tree in /usr/lib