vincent_kwinsey wrote:

>
> I have the similar error message - but for page type 5. The problem
> is - that after this usually the following other error messages start
> to appear:
> 'internal gds software consistency check (cannt continue after
> bugcheck)'
>
> So - it can be understood quite well, because - if there is no chance
> to write the required information, the DB ond-disk information can
> get corrupted.

In theory, the internal data structures are deadlock free - at least
that's the design goal. The comment I quoted is from the intersection
of theory and practice ... the lock manager has detected a deadlock
where there should never bee a deadlock. If the problem can be
reproduced, perhaps the design problem can be found...


Regards,


Ann