Обсуждение: Postgre cant start
Hello,
If you can be of any help… please feel free.
PostgreSQL V7.1
==================================================
[postgres@x bin]$ postgres start
DEBUG: database system was interrupted being in recovery at 2008-04-01 00:31:20 CST
This propably means that some data blocks are corrupted
and you will have to use last backup for recovery.
DEBUG: CheckPoint record at (6, 2667839868)
DEBUG: Redo record at (6, 2667793068); Undo record at (0, 0); Shutdown FALSE
DEBUG: NextTransactionId: 58839239; NextOid: 18823136
DEBUG: database system was not properly shut down; automatic recovery in progress...
DEBUG: redo starts at (6, 2667793068)
FATAL 2: cannot read block 1802 of pg_log: Success
[root@x log]# cat boot.log |grep postmaster
Apr 1 04:48:42 mail1 postgres: postmaster shutdown failed
Apr 1 04:49:38 mail1 postgres: postmaster shutdown failed
Apr 1 04:51:40 mail1 postgres: postmaster shutdown failed
Mar 31 22:59:14 mail1 postgres: postmaster shutdown failed
Mar 31 22:59:32 mail1 postgres: postmaster shutdown failed
Mar 31 23:01:45 mail1 postgres: postmaster shutdown failed
Mar 31 23:38:10 mail1 postgres: postmaster shutdown failed
Regards
Olivier Bessière
Network and System Engineer
Think green before printing!
__________ Information from ESET NOD32 Antivirus, version of virus signature database 2992 (20080401) __________
The message was checked by ESET NOD32 Antivirus.
http://www.eset.com
Olivier Bessière <obessiere@boomerangpharma.com> writes: > PostgreSQL V7.1 Egad. You really, really, really need to update forthwith. The list of bugs that have been fixed since 7.1 would curl your toes. http://developer.postgresql.org/pgdocs/postgres/release.html > FATAL 2: cannot read block 1802 of pg_log: Success This suggests that pg_log has been truncated. Try something like dd bs=8192 count=1 seek=1802 if=/dev/zero of=$PGDATA/global/1269 and try to start the postmaster. If it then complains about a higher block number, add that one in the same way, and repeat. I don't have a whole lot of confidence that this will get you a self-consistent database, but maybe you will be able to extract some data. Dump, compare against your last backup (you do have one I trust), see what data you believe. For heaven's sake, update! regards, tom lane
Hello, Actually... You're command line just fully fixed my problem. Postmaster is starting like he never got any problem. Thank you so much for this helpful answer. How can I thank you ? Are you traveling to France soon ? Regards /Olivier -----Original Message----- From: Tom Lane [mailto:tgl@sss.pgh.pa.us] Sent: mardi 1 avril 2008 20:50 To: Olivier Bessière Cc: pgsql-bugs@postgresql.org Subject: Re: [BUGS] Postgre cant start Olivier Bessière <obessiere@boomerangpharma.com> writes: > PostgreSQL V7.1 Egad. You really, really, really need to update forthwith. The list of bugs that have been fixed since 7.1 would curl your toes. http://developer.postgresql.org/pgdocs/postgres/release.html > FATAL 2: cannot read block 1802 of pg_log: Success This suggests that pg_log has been truncated. Try something like dd bs=8192 count=1 seek=1802 if=/dev/zero of=$PGDATA/global/1269 and try to start the postmaster. If it then complains about a higher block number, add that one in the same way, and repeat. I don't have a whole lot of confidence that this will get you a self-consistent database, but maybe you will be able to extract some data. Dump, compare against your last backup (you do have one I trust), see what data you believe. For heaven's sake, update! regards, tom lane __________ Information from ESET NOD32 Antivirus, version of virus signature database 2992 (20080401) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com __________ Information from ESET NOD32 Antivirus, version of virus signature database 2994 (20080402) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com