Обсуждение: Postmaster will not restart

mark hewitt (m.hewitt@computer.org) reports a bug with a severity of 1
The lower the number the more severe it is.

Short Description
Postmaster will not restart

Long Description
Afetr several very happy years of using PostgreSQL, I finally
have a serious problem.
As far as I know, the postmaster terminated correctly, but now
when I restart it I have:

DEBUG:  database system was shut down at 2001-07-30 21:17:17 BST
DEBUG:  CheckPoint record at (0, 214140268)
DEBUG:  Redo record at (0, 214140268); Undo record at (0, 0); Shutdown TRUE
DEBUG:  NextTransactionId: 221440; NextOid: 617393
FATAL 2:  ZeroFill(/work/database/postgres/pg_xlog/xlogtemp.449)
failed: No such file or directory

Is there any way back from here?

Indeed, the xlogtemp file does not exist, and that directory contains:

-rw-------    1 postgres db       16777216 Jul 30 21:17 000000000000000C

My platform is PostgreSQL 7.1.2 on Linux x86 (2.4.7 kernel) with
glibc 2.2.3.

#!/mjh

Sample Code


No file was uploaded with this report

pgsql-bugs@postgresql.org writes:
> As far as I know, the postmaster terminated correctly, but now
> when I restart it I have:

> DEBUG:  database system was shut down at 2001-07-30 21:17:17 BST
> DEBUG:  CheckPoint record at (0, 214140268)
> DEBUG:  Redo record at (0, 214140268); Undo record at (0, 0); Shutdown TRUE
> DEBUG:  NextTransactionId: 221440; NextOid: 617393
> FATAL 2:  ZeroFill(/work/database/postgres/pg_xlog/xlogtemp.449)
> failed: No such file or directory

Looks to me like you have insufficient free space on your disk --- the
code is trying to create another 16-MB WAL segment, and is failing,
no doubt for lack of free space.  (Yes, the quoted kernel error message
is incorrect ... that problem is patched in current sources.)

If you do have more than 16MB free on your WAL disk, then there's
another problem...

            regards, tom lane

On 30-Jul-2001 Tom Lane wrote:

>> DEBUG:  database system was shut down at 2001-07-30 21:17:17 BST
>> DEBUG:  CheckPoint record at (0, 214140268)
>> DEBUG:  Redo record at (0, 214140268); Undo record at (0, 0); Shutdown
>> TRUE
>> DEBUG:  NextTransactionId: 221440; NextOid: 617393
>> FATAL 2:  ZeroFill(/work/database/postgres/pg_xlog/xlogtemp.449)
>> failed: No such file or directory
>
> Looks to me like you have insufficient free space on your disk --- the
> code is trying to create another 16-MB WAL segment, and is failing,
> no doubt for lack of free space.  (Yes, the quoted kernel error message
> is incorrect ... that problem is patched in current sources.)

Embarassing, isn't it!  The disc space was being consumed by a transient
log file elsewhere on the same drive.  Thanks for the humane treatment
of such a mundae problem!

--
Mark J Hewitt at home           E-Mail: Mark J Hewitt <m.hewitt@computer.org>
Date: 31-Jul-2001  Time: 00:09:44

Mark J Hewitt <m.hewitt@computer.org> writes:
> Thanks for the humane treatment of such a mundae problem!

Well, it *is* a bug, albeit of the "misleading error message" variety
rather than anything more critical.  If you are interested in applying
a source patch to fix that problem, see

http://www.ca.postgresql.org/mhonarc/pgsql-patches/2001-06/msg00061.html

            regards, tom lane