Failure to extend rollback segment N because of 1000 condition
Date: Fri, 14 May 2010 14:52:28 -0700 (PDT)
Message-ID: <1da7363a-cc2e-416d-b38e-a3af1d45447a_at_h37g2000pra.googlegroups.com>
I got strange error message in alert.log for 9.2.0.7 database (64-bit) on AIX 5.3
ARC0: Completed archiving log 4 thread 1 sequence 123305
Fri May 14 21:05:55 2010
Failure to extend rollback segment 15 because of 1000 conditionFULL
status of rollback segment 15 set.
Fri May 14 21:05:59 2010
I am familiar with 30036 error, but not with 1000. In fact error 1000 has nothing to do with rollback/undo
oracle> oerr ora 30036
30036, 00000, "unable to extend segment by %s in undo tablespace '%s'"
// *Cause: the specified undo tablespace has no more space
available.
// *Action: Add more space to the undo tablespace before retrying
// the operation. An alternative is to wait until active // transactions to commit.
oracle> oerr ora 1000
01000, 00000, "maximum open cursors exceeded"
// *Cause:
// *Action:
The database is configured with automatic undo management. This is
batch processing system, undo generation is stable at approximately
60,000 blocks per 10 min interval (as per V$UNDOSTAT), or 360,000
blocks per hour. Undo retention is set to 28,000 seconds, or 8 hours.
To keep 8 hours of undo we need approximately 3 million blocks
(2,880,000), or 24 GB as database uses 8K blocks. Undo tablespace is
60 GB, approx 50% free.
So everything seems to be configured correctly, why the error?
Received on Fri May 14 2010 - 16:52:28 CDT