Re: "PANIC: could not open critical system index 2662" - twice

From: Evgeny Morozov <postgresql3(at)realityexists(dot)net>
To: Thomas Munro <thomas(dot)munro(at)gmail(dot)com>
Cc: PostgreSQL General <pgsql-general(at)postgresql(dot)org>
Subject: Re: "PANIC: could not open critical system index 2662" - twice
Date: 2023-05-06 09:58:01
Message-ID: 01020187f07f341f-d1593cf8-4fc7-4186-b70c-0c8862a2aa32-000000@eu-west-1.amazonses.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 6/05/2023 1:06 am, Thomas Munro wrote:
> Next can you share the file base/1414389/2662? ("5" was from the wrong
> database.)

Right - I should have realised that! base/1414389/2662 is indeed all
nulls, 32KB of them. I included the file anyway in
https://objective.realityexists.net/temp/pgstuff2.zip

> Try something like:
> pg_waldump -R 1663/1414389/2662 -F main 000000010000000000000001
> 000000010000000000000007
>
> ... but change that to the range of files you have in your pg_wal.

The PG server had only 2 files left from today, and pg_waldumping them
gave this error:

pg_waldump -R 1663/1414389/2662 -F main 00000001000000650000005E
00000001000000650000005F
pg_waldump: error: error in WAL record at 65/5F629838: invalid record
length at 65/5F62A1E0: wanted 24, got 0

Not sure if that's something to worry about or not!

Then I realised we're actually archiving our WAL files with pgBackRest,
retrieved the WAL files for the time the DB was created and used (~12:39
UTC on 2023-05-02) and re-ran pg_waldump on those.

pg_waldump -R 1663/1414389/2662 -F main 000000010000005B00000000
000000010000005B0000000F

rmgr: XLOG        len (rec/tot):     51/   108, tx:     242382, lsn:
5B/0222BC68, prev 5B/0222BC38, desc: FPI , blkref #0: rel
1663/1414389/2662 blk 0 FPW
rmgr: XLOG        len (rec/tot):     51/  3224, tx:     242382, lsn:
5B/0222BCD8, prev 5B/0222BC68, desc: FPI , blkref #0: rel
1663/1414389/2662 blk 1 FPW
rmgr: XLOG        len (rec/tot):     51/  1544, tx:     242382, lsn:
5B/0222C988, prev 5B/0222BCD8, desc: FPI , blkref #0: rel
1663/1414389/2662 blk 2 FPW
rmgr: XLOG        len (rec/tot):     49/   121, tx:     242382, lsn:
5B/0222CF90, prev 5B/0222C988, desc: FPI , blkref #0: rel
1663/1414389/2662 blk 3 FPW
rmgr: Btree       len (rec/tot):     64/    64, tx:     242384, lsn:
5B/02321870, prev 5B/023217A0, desc: INSERT_LEAF off 132, blkref #0: rel
1663/1414389/2662 blk 2
rmgr: Btree       len (rec/tot):     64/    64, tx:     242384, lsn:
5B/02322640, prev 5B/02322570, desc: INSERT_LEAF off 133, blkref #0: rel
1663/1414389/2662 blk 2
... (many more entries like that)
pg_waldump: error: error in WAL record at 5B/FFFFF38: missing contrecord
at 5B/FFFFF70

The full output is also in
https://objective.realityexists.net/temp/pgstuff2.zip

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Thomas Munro 2023-05-06 10:34:26 Re: "PANIC: could not open critical system index 2662" - twice
Previous Message Peter J. Holzer 2023-05-06 07:46:25 Re: Death postgres