Обсуждение: failed to re-find parent key in index "22723655" for split pages 370836/370837

Поиск
Список
Период
Сортировка

failed to re-find parent key in index "22723655" for split pages 370836/370837

От
Anderson Valadares
Дата:
Hello
 
I am facing problems starting the database. Following is the error message. Can anyone help me?

Parameters

synchronous_commit = off
archive_mode = on
sync = on


Postgres 8.4.5
Linux 2.6.18-194.3.1.el5 # 1 SMP Thu May 13 13:08:30 EDT 2010 x86_64 x86_64 x86_64 GNU / Linux

2010-11-29 18:12:20 BRST [] - [5011]: [1-1] LOG:  00000: database system was interrupted while in recovery at 2010-11-29 18:09:31 BRST
2010-11-29 18:12:20 BRST [] - [5011]: [2-1] HINT:  This probably means that some data is corrupted and you will have to use the last backup for recovery.
2010-11-29 18:12:20 BRST [] - [5011]: [3-1] LOCATION:  StartupXLOG, xlog.c:5265
2010-11-29 18:12:20 BRST [] - [5011]: [4-1] LOG:  00000: database system was not properly shut down; automatic recovery in progress
2010-11-29 18:12:20 BRST [] - [5011]: [5-1] LOCATION:  StartupXLOG, xlog.c:5449
2010-11-29 18:12:20 BRST [] - [5011]: [6-1] LOG:  00000: redo starts at 270/55E04AE8
2010-11-29 18:12:20 BRST [] - [5011]: [7-1] LOCATION:  StartupXLOG, xlog.c:5532
2010-11-29 18:12:41 BRST [] - [5011]: [8-1] LOG:  58P01: could not open file "pg_xlog/0000000100000270000000CC" (log file 624, segment 204): No such file or directory
2010-11-29 18:12:41 BRST [] - [5011]: [9-1] LOCATION:  XLogFileRead, xlog.c:2600
2010-11-29 18:12:41 BRST [] - [5011]: [10-1] LOG:  00000: redo done at 270/CBFFE940
2010-11-29 18:12:41 BRST [] - [5011]: [11-1] LOCATION:  StartupXLOG, xlog.c:5664
2010-11-29 18:12:41 BRST [] - [5011]: [12-1] LOG:  00000: last completed transaction was at log time 2010-11-26 01:55:01.259996-02
2010-11-29 18:12:41 BRST [] - [5011]: [13-1] LOCATION:  StartupXLOG, xlog.c:5668
2010-11-29 18:12:41 BRST [] - [5011]: [14-1] FATAL:  XX000: failed to re-find parent key in index "22723655" for split pages 370836/370837
2010-11-29 18:12:41 BRST [] - [5011]: [15-1] LOCATION:  _bt_insert_parent, nbtinsert.c:1578
2010-11-29 18:12:41 BRST [] - [5009]: [1-1] LOG:  00000: startup process (PID 5011) exited with exit code 1
2010-11-29 18:12:41 BRST [] - [5009]: [2-1] LOCATION:  LogChildExit, postmaster.c:2720
2010-11-29 18:12:41 BRST [] - [5009]: [3-1] LOG:  00000: aborting startup due to startup process failure
2010-11-29 18:12:41 BRST [] - [5009]: [4-1] LOCATION:  reaper, postmaster.c:2261

Re: failed to re-find parent key in index "22723655" for split pages 370836/370837

От
Tom Lane
Дата:
Anderson Valadares <andervalbh@gmail.com> writes:
>  I am facing problems starting the database. Following is the error message.
> Can anyone help me?

> 2010-11-29 18:12:41 BRST [] - [5011]: [14-1] FATAL:  XX000: failed to
> re-find parent key in index "22723655" for split pages 370836/370837

The only way you're going to get out of that is with pg_resetxlog,
which is a fairly blunt instrument :-(.  It's possible that you will
have database corruption afterwards.  I'd recommend doing a dump and
reload to see if any problems appear.  It'd also be a good idea to save
a tarball copy of the $PGDATA file tree beforehand, so at least you can
get back to where you are in case things turn into a complete mess.

Can you tell us anything about how you got into this state?

            regards, tom lane

Re: failed to re-find parent key in index "22723655" for split pages 370836/370837

От
Anderson Valadares
Дата:
2010/11/29 Tom Lane <tgl@sss.pgh.pa.us>
Anderson Valadares <andervalbh@gmail.com> writes:
>  I am facing problems starting the database. Following is the error message.
> Can anyone help me?

> 2010-11-29 18:12:41 BRST [] - [5011]: [14-1] FATAL:  XX000: failed to
> re-find parent key in index "22723655" for split pages 370836/370837

The only way you're going to get out of that is with pg_resetxlog,
which is a fairly blunt instrument :-(.  It's possible that you will
have database corruption afterwards.  I'd recommend doing a dump and
reload to see if any problems appear.  It'd also be a good idea to save
a tarball copy of the $PGDATA file tree beforehand, so at least you can
get back to where you are in case things turn into a complete mess.

Can you tell us anything about how you got into this state?

                       regards, tom lane

Thanks Tom,
is strange because the database was shut down normally. I wanted to understand why this error. I noticed that this table has an index GIST.

27/11/2010 13:16:49 PDT [] - [13019]: [3-1] LOG: received smart shutdown request
27/11/2010 13:16:49 PDT [] - [3891]: [1-1] FATAL: terminating connection due to administrator command
27/11/2010 13:16:49 PDT [] - [28117]: [1-1] FATAL: terminating connection due to administrator command
27/11/2010 13:16:49 PDT [] - [25348]: [8597-1] FATAL: terminating connection due to administrator command
27/11/2010 13:16:49 EDT [- [3557]: [8-1] FATAL: terminating connection due to administrator command
27/11/2010 13:16:49 PDT [] - [26900]: [1-1] FATAL: terminating connection due to administrator command
27/11/2010 13:16:49 PDT [] - [27051]: [1-1] FATAL: terminating connection due to administrator command