Re: Intentando recuperar datos: pgsql 8.0.1 Windows

On 9/25/07, Luis Perruca <l(dot)perruca(at)gmail(dot)com> wrote:
> Hola al grupo, y perdonad por la tardanza, pero estuve de vacaciones ;)
>
> No me resigno a perder los datos, así que voy a intentar aportar toda la
> información que pueda:
>

bueno, te tengo malas noticias:

1)
2007-08-16 14:56:02 FATAL: could not open relation
1663/17254/1233124: No such file or directory
2007-08-16 14:56:02 FATAL: could not open relation
1663/17254/1233124: No such file or directory

Esto es que no encuentra un objeto (se fue, caput, se esfumo)

2)
2007-08-16 14:52:21 LOG: database system was interrupted while in
recovery at 2007-08-16 14:51:26 Hora de verano romance
2007-08-16 14:52:21 HINT: This probably means that some data is
corrupted and you will have to use the last backup for recovery.

data corrupted!!! que querra decir?

3)
2007-08-16 14:52:21 LOG: checkpoint record is at 0/20343D8
2007-08-16 14:52:21 LOG: redo record is at 0/20343D8; undo record is
at 0/0; shutdown TRUE
2007-08-16 14:52:21 LOG: next transaction ID: 5333; next OID: 33637
2007-08-16 14:52:21 LOG: database system was not properly shut down;
automatic recovery in progress
2007-08-16 14:52:21 LOG: redo starts at 0/2034418
2007-08-16 14:52:21 PANIC: btree_insert_redo: block unfound

esto no suena bien!!!

4)
2007-08-16 14:45:27 ERROR: xlog flush request 6A/8B91AEA0 is not
satisfied --- flushed only to 0/20BDF30
2007-08-16 14:45:27 CONTEXT: writing block 1170 of relation 1663/17254/17586
2007-08-16 14:45:27 WARNING: could not write block 1170 of 1663/17254/17586
2007-08-16 14:45:27 DETAIL: Multiple failures --- write error may be permanent.
2007-08-16 14:45:27 ERROR: FlushRelationBuffers("pg_statistic", 43):
block 2 is referenced (private 0, global 1)
2007-08-16 14:45:27 PANIC: cannot abort transaction 5783, it was
already committed

en definitiva me parece que deberias cambiar tu disco... y como dice
uno de los mensajes del log (deberias restaurar el backup mas
reciente, y no se refiere a la misma carpeta corrupta que tienes)...

la unica otra opcion que se me ocurre (que no implique un disco
dañado) es que tengas: discos ide; fsync = off; y un apagon en medio
proceso de algo... bueno, es solo una idea, no se si esa combinacion
podria causar el problema que describen los logs...

--
Atentamente,
Jaime Casanova

"Programming today is a race between software engineers striving to
build bigger and better idiot-proof programs and the universe trying
to produce bigger and better idiots.
So far, the universe is winning."
Richard Cook