Re: Unable to dump database/table

From: Andre Maasikas <andre(at)abs(dot)ee>
To: Krok <krok(at)void(dot)ru>
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: Unable to dump database/table
Date: 2004-10-11 17:21:21
Message-ID: Pine.LNX.4.50.0410112011150.2890-100000@abbs03xe3.corp.abs.ee
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On Mon, 11 Oct 2004, Krok wrote:

> void=> select max(length(passwd)) from users;
> max
> -----------
> 536870919
> (1 row)

....
>
> But how can this happen, that varchar(255) field became broken ?
>
According to my caclulator, in hex this is 0x20000007
so there's likely a bit flip if the length is actually stored
so ( i don't know the internals yet )

I'd be worried about hardware and more importantly about data
consistency at this point. I would suggest running manufacturer's
test suites or something similar on this machine.

(memtest on x86 would be good, copying some .iso's around and
checksumming them in between was a good way for me to ask
a replacement machine once)

Andre

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Reuven M. Lerner 2004-10-11 18:00:19 Fatal "make check" bug with 8.0 beta 3 under Mac OS 10.3
Previous Message Tom Lane 2004-10-11 16:35:35 Re: Why is the index not created in the tablespace delivered in the create-index-command ?