As it turns out, the problem is only when you migrate the catalog using export import. What you should do is increse the controlfile record keep time, sync it, connect to the new catalog and regiser the DB. you can reduce the cf_record_keep_time after the first resync with the new catalog. I'd avoid using local disks if at all possible, but I don't think the problem was with the hardware. We chose to downgrade the catalog to 10.2.0.4.0, so I don't really know what happened with the bug. cheers. Alan.- On Tue, Jul 13, 2010 at 12:23 PM, David Robillard <david.robillard@xxxxxxxxx > wrote: > Hi Alan, > > > They finally opened a new bug for us, 9630883 and it's being worked on by > Development. > > > > I know the catalog is a rather simple thing, but even the simplest things > > have bugs... and we happened to hit one. > > I've just installed a machine with 11gR2 to replace our current 9iR2 > RMAN catalog. So I'm very interested by this bug you found. > Would you care to give me more information about this problem? > > If it's relevant, the machine is running RedHat Linux 32 bit with the > catalog on local disks. > > Thanks, > > David >