Re: [ADMIN] Problem with pg_dump

2009/1/13 <tyrrill_ed(at)emc(dot)com>:
> Hi All,
>
>
>
> I am a developer of a product that uses a postgresql database (currently
> version 8.2.3.1). We dump the database using pg_dumpall. We are finding
> data corruption in the dump files about twice a month with a few thousand
> installations. I have been able to track down the data corruption to the
> original dump file created by pg_dumpall. In every case I've seen of the
> corruption 1 or 2 characters are missing. The problem exhibits itself when
> loading the dump files. Here is an example where two lines were combined:
>
>

you try to use pg_dump instead (only the DB you want)?
what fs you are using? (i think is ntfs)
Integrity check?
Force the enconding of dump to the same as OS.

I don't know about a bug in these version, specially in terms of data.

>
> ERROR: extra data after last expected column
>
> CONTEXT: COPY ds_targets, line 42: "1180635517879 3001
> C:/SQLBackup/1180635517879 3001 C:/System Volume Information/"
>
>
>
> The file bad row was:
>
>
>
> 1180635517879 3001 C:/SQLBackup/1180635517879 3001 C:/System
> Volume Information/
>
>
>
> It should have been two rows as follows:
>
>
>
> 1180635517879 3001 C:/SQLBackup/
>
> 1180635517879 3001 C:/System Volume Information/
>
>
>
> Is this something that is perhaps fixed in postgresql 8.3? Any assistance
> in resolving this problem would be appreciated.
>
>
>
> Thanks,
>
> Ed
>
>
>
>

--
Emanuel Calvo Franco
ArPUG / AOSUG Member
Postgresql Support & Admin