Обсуждение: Re: [Support i5GKDLjR008723] PGRES_FATAL_ERROR: out of free buffers:

Hi, Jochen.

> However, we frequently observe the following message in the log files:
>    PGRES\_FATAL_ERROR : ERROR:  out of free buffers: time to abort!
> Any hints what's going on? Suggestion?

Two things might help:

1) Upgrade to the latest CanIt (2.1a), because earlier versions seemed
to trigger this bug in PostgreSQL.

2) Upgrade to the latest version of PostgreSQL (7.4.3 as of this writing.)
If you upgrade from PostgreSQL earlier than 7.4, be aware that you must
dump all your databases with the old version *before* starting the upgrade.

Regards,

David.

"David F. Skoll" <dfs@roaringpenguin.com> writes:
> Hi, Jochen.
>> However, we frequently observe the following message in the log files:
>> PGRES\_FATAL_ERROR : ERROR:  out of free buffers: time to abort!
>> Any hints what's going on? Suggestion?

IIRC, there was a problem in 7.3 - 7.3.2 that tended to provoke this
failure.  It's repaired in 7.3.3 and later.  If you want to stay in
the 7.3 series rather than upgrade to 7.4, I would *strongly* recommend
moving to 7.3.6.

            regards, tom lane

Jochen Wiedmann <jochen.wiedmann@freenet.de> writes:
> thanks for the hints. Following you, I have upgraded to 7.3.4 (the last
> version for which an RPM by Red Hat is provided) and the problem seems
> gone now.

?? I thought I'd pushed out a 7.3.6 RPM for RHEL3.  Where exactly are
you finding 7.3.4 as still the latest?

There are several nasty problems in 7.3.4, so I don't want to see you
settling for that ...

            regards, tom lane

Tom, David,

thanks for the hints. Following you, I have upgraded to 7.3.4 (the last
version for which an RPM by Red Hat is provided) and the problem seems
gone now.


Regards,

Jochen


On Fr, 2004-06-18 at 17:22, Tom Lane wrote:
> "David F. Skoll" <dfs@roaringpenguin.com> writes:
> > Hi, Jochen.
> >> However, we frequently observe the following message in the log files:
> >> PGRES\_FATAL_ERROR : ERROR:  out of free buffers: time to abort!
> >> Any hints what's going on? Suggestion?
>
> IIRC, there was a problem in 7.3 - 7.3.2 that tended to provoke this
> failure.  It's repaired in 7.3.3 and later.  If you want to stay in
> the 7.3 series rather than upgrade to 7.4, I would *strongly* recommend
> moving to 7.3.6.

Tom Lane wrote:

> ?? I thought I'd pushed out a 7.3.6 RPM for RHEL3.  Where exactly are
> you finding 7.3.4 as still the latest?
>
> There are several nasty problems in 7.3.4, so I don't want to see you
> settling for that ...

I wasn't aware, that RHEL3 RPM's can be used on RH9. Following your mail, I
   gave it a try, and it works.


Regards,

Jochen