Re: bug in 6.0.36.0

In article <1993Jun16.155346_at_abds7.aberdeen.chevron.com> epeas_at_aberdeen.chevron.com writes:
>
>Can any one enlighten me as to the symptoms of corruption as caused by bug
>130548 under 6.0.36.0?
>

As I understand the problem it is associated with one oracle user not properly detecting a lock on a datablock. The risk then would be that two (or more) users could change a datablock at the same time. So your corruption would come from the begining of the datablock not matching the end.

The best way of finding a bad block with data in it would be an export. If you had a bad block exp would tell you that it had an export error and was skipping that table. Of course this would not do any thing for a mangled extent that the table thought was empty but might try to use in the future.

The support letter I read several months ago mentioned that this was associated only with the parrallel server option. Give the type of error this makes sense but Oracle has made too big a deal of this over too many platforms for this to be true.

-- 
Regards, 

Lee E. Parsons                  		Baker-Huges Inteq, Inc
Oracle Database Administrator 			lparsons_at_exlog.com