> I've also had problems with long running sweep(afterI had deleted
> about 1 million records from one table). Eventually following things
> sort problem out:
> 1.gbak -b -g (inhibit garbage collection)
> 2.gbak -r (restore)
> After that sweep took only 1 minute. I think that indices were reason
> for long sweep.

Was that after the sweep finished?
It has been running for about 133 hours, and is still going.
Should i just stop it and do a gbak -b -g and then a gbak?

Thanks