Goodness. You dropped a data file out of the SYSTEM tablespace? Or are you writing that you dropped the tablespace containing the object referenced by the corrupt I_OBJ2 index? (All the indexes of sys.obj$ are born in SYSTEM unless you have an altered sql.bsq file.) If that is the case, did the offline drop clean up the entries in I_OBJ2? Then I'm surprised just dropping the offending referenced object didn't do the trick. If no clean up took place, then I would think you still have a corrupt block in I_OBJ2. At this point I'm mostly just curious about exactly what did transpire, since I don't understand. Regards, mwf -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Grzegorz Goryszewski Sent: Saturday, November 26, 2011 10:29 AM To: mwf@xxxxxxxx Cc: oracle-l@xxxxxxxxxxxxx Subject: Re: value from index block or table block On 2011-11-26 14:53, Mark W. Farnham wrote: > Let me get this straight: You are experiencing the return of incorrect > values (meaning either a bug or a corruption) and Oracle Support will > not authorize your use of setting an event either at the session or > system level? > > Hi Mark, sorry but looks like I was no precise enough . We provided 10046 ,and that confirmed corruption in I_OBJ2 index (if memory serves me right name) but as a solution oracle said please do exp/imp of your database . Looks cool but our db is 15TB in size :) . The problem is You cant rebuild that index (cant remember exac ora err for that ) in normal way, You have to open database in upgrade/migrate mode and with some event try to rebuild but its not supported as far as I know . Finally we did drop offline for data file with that object and physically removed it . Only problem is rman warning about that unrecoverable tablespace . Regards GregG -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l