Oracle Community Network
We have oracle 11gr2 database on linux 5.5. for this database undo tablespace is 7GB.
Frequently we are receiving alerts for this table space like 100% full.
I tried to figure out the optimum size for this undo tablespace and found one formula for this: (undo blocks generating per second* undo retention * undo tablespace block size)
I took blocks generating per second from v$undostat view. when i calculated, strange it give the value only 1.2GB. that mean the optimum size is 1.2GB
but i allocated 7gb of undo tablespace but still receiving the alerts.
is there any reason for this kind of alerts and am i following the right formula.
Please help to fix this.
disable or turn off the Chicken Little nag-o-gram generator that issues false positive alerts.
You only need to fix real errors which contain error number & message.
A "full" undo tablespace is normal - see MOS Docs 413732.1 and 1335964.1