This usually implies that a query has run out of rollback records needed for a consistent read, as they are overwritten by other writers to the database. This would make the result of the query unreliable. You probably need to increase the size of your rollback segments (or use a special 'bulk' one) for this query if the error happens consistently. HTH.
Ooops, there are several reasons for this error other than the obvious. If you still encounter this problem after increasing your rollback segments, then there must be other causes.
Here's a list of the causes of this error.
Cause Description
----- -----------
1 A long query takes place at the same time as an update
2 Read consistent view during "insert"
3 Delayed block cleanout on old committed updates
4 Fetch across commit
5 Rollback segment corruption
You can search technet for white papers on this one.
Robbie
"The rule is, not to besiege walled cities if it can possibly be avoided" -- Art of War
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.