Thinking operationally, there are always a myriad number of methods to track usage and report on UNDO statistics. However, if attempts have been made to increase these resources, and problems still remain, I have never seen a time when such issues can be resolved WITHOUT looking into the root causes.
Most of the time, a particular query or combination of queries will be the issue. Rather than throwing disk resources at the problem, it is usually necessary to first determine what is running when it occurs, and find the source or origination of these jobs.
Inevitably, a particular query (or sometimes the same query running concurrently) will be the cause and will have to be tuned. We have canned queries which will report on ALL tablespace full conditions which I can supply, but I really think you need to investigate the cause before heading in that direction...
I have seen several time ORA-1555 error in few of the Prod Database. SOme of them have very big UNDO tablespaces. All are 9i Databases. What are the other option that need to be checked for such issue beside increasing the space as some of the database has UNDO tablespace around 50-150G
Is it good to create new undo tablespace and make it default in such scenario and later on dropped the old. Do somebody has set of UNDO checking query that can be ran at some particular point to really see if that big UNDO are really been used so that it can be resized.
I am thankful for any suggestion for monitoring such ORA-1555 as well as query to check UNDO segment space.
Finding fabulous fares is fun.Let Yahoo! FareChase search your favorite travel sites
to find flight and hotel bargains.