ERROR: unexpected data beyond EOF in block of relation "RelationName"

Started by Marcin Gonover 16 years ago3 messagesgeneral
Jump to latest
#1Marcin Gon
marcin77@ymail.com

Hi,

I'm getting the following error from my Postgres database while inserting:

ERROR: unexpected data beyond EOF in block of relation "RelationName".

My configuration is:
openSUSE 11.1
Linux linux-wsr1 2.6.27.23-0.1-default #1 SMP 2009-05-26 17:02:05 -0400 x86_64 x86_64 x86_64 GNU/Linux
Postgres 8.3.7

Can anybody advise how to solve this problem?

Thanks!
Marcin

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: Marcin Gon (#1)
Re: ERROR: unexpected data beyond EOF in block of relation "RelationName"

Marcin Gon <marcin77@ymail.com> writes:

I'm getting the following error from my Postgres database while inserting:

ERROR: unexpected data beyond EOF in block of relation "RelationName".

You omitted the HINT that says this has only been known to occur in
connection with buggy kernels. I see you're running a fairly new
kernel, but it might still pay to inquire whether it's subject to the
wrong-lseek-result race condition that was around ahile back.

As far as getting out of the immediate problem is concerned, I think
restarting the postmaster ought to fix it.

regards, tom lane

#3Marcin Gon
marcin77@ymail.com
In reply to: Tom Lane (#2)
Re: ERROR: unexpected data beyond EOF in block of relation "RelationName"

Hi,

Thanks for that. The question for me is how to find an appropriate OS patch for this? Is there a list of required patches for this Postgres release on SuSE?

Regards,
Marcin

--- On Thu, 23/7/09, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Show quoted text

From: Tom Lane <tgl@sss.pgh.pa.us>
Subject: Re: [GENERAL] ERROR: unexpected data beyond EOF in block of relation "RelationName"
To: "Marcin Gon" <marcin77@ymail.com>
Cc: pgsql-general@postgresql.org
Date: Thursday, 23 July, 2009, 2:48 PM
Marcin Gon <marcin77@ymail.com>
writes:

I'm getting the following error from my Postgres

database while inserting:

ERROR: unexpected data beyond EOF in block of relation

"RelationName".

You omitted the HINT that says this has only been known to
occur in
connection with buggy kernels.  I see you're running a
fairly new
kernel, but it might still pay to inquire whether it's
subject to the
wrong-lseek-result race condition that was around ahile
back.

As far as getting out of the immediate problem is
concerned, I think
restarting the postmaster ought to fix it.

           
regards, tom lane