Just to to finish this thread:
We had a corrupted memory bank in the development server, and that
was, for some reason corrupting the data that got written to disk.
All is good now. Sorry for the noise.
El día 23 de marzo de 2011 16:28, Martín Marqués
<martin.marques@gmail.com> escribió:
> 2011/3/23 Vick Khera <vivek@khera.org>:
>> 2011/3/23 Martín Marqués <martin.marques@gmail.com>:
>>> The second error is very curious, as it looks like pg_dump changed 1
>>> for a 'q' in an integer field:
>>>
>>> psql:siprebi-bu.sql:2219245: ERROR: la sintaxis de entrada no es
>>> válida para integer: «q721695»
>>> CONTEXT: COPY objeto_datos_rep, línea 991894, columna codigo: «q721695»
>>>
>>
>> That's a single-bit error. I'm voting hardware problems. Does your
>> memory have ECC? How reliable is your disk?
>>
>
> The production server is a Compaq Prolaint, with ECC memory and RAID
> by hardware.
>
> How can I discard, or affirm that we are dealing with hardware problems?
>
> --
> Martín Marqués
> select 'martin.marques' || '@' || 'gmail.com'
> DBA, Programador, Administrador
>
--
Martín Marqués
select 'martin.marques' || '@' || 'gmail.com'
DBA, Programador, Administrador