Subject | Re: 2 identical record inserts not stopped by UNIQUE constraint |
---|---|
Author | Stan |
Post date | 2006-10-25T16:01:39Z |
Adam,
I suspect that this issue is a result of the corruption as well.
No, I have no idea what caused the corruption, The machine
did not reboot or crash. The FB server did not die.
the Firebird.log only contains a bunch of 10061 socket errors
which seem to be "normal" during high load.
I guess that it is not out of the question that this machine has bad
RAM, maybe... No other issues on this machine so far.
If this happens again, I will assume that it is a hardware issue that
caused the corruption since I have not seen this happen on other test
machines.
thanks,
stan
I suspect that this issue is a result of the corruption as well.
No, I have no idea what caused the corruption, The machine
did not reboot or crash. The FB server did not die.
the Firebird.log only contains a bunch of 10061 socket errors
which seem to be "normal" during high load.
I guess that it is not out of the question that this machine has bad
RAM, maybe... No other issues on this machine so far.
If this happens again, I will assume that it is a hardware issue that
caused the corruption since I have not seen this happen on other test
machines.
thanks,
stan
--- In firebird-support@yahoogroups.com, "Adam" <s3057043@...> wrote:
>
>
> > Number of index page errors : 123
>
> Stan,
>
> This looks suspicious to me. A unique constraint simply checks the
> underlying index to determine whether it has been violated. If you are
> having corruption in index pages, all bets are off.
>
> Any hints on what may have caused the corruption?
>
> Adam
>