Today's duty in case of really many, mirror database a large number of rollback, can not be started normally; A database inexplicable downtime; also within a month the second encountered a ORA-600, is it time to burn incense.
Or simply record this ORA-600.
OS: aix 5.3
DB: 10.2.0.4
After receiving a warning of a large number of transaction congestion in a database, the database was found to have a large number of enq: US-contention waits after logging on through a remote VPN. Check the lock situation and there are a large number of US locks.
View the alarm log and find that the alarm log is always reported
ORA-00600: internal error code, arguments: [4193], [41520], [41523], [], [], [], [], [], []
.
See materials to learn
4193: indicates that undo and redo are inconsistent (Arg [a] Undo record seq number, Arg [B] Redo record seq number );
ORA-600 [4193] WhenTrying To Open The Database [ID 763566.1]
Try to recreate the undo tablespace. As expected, there will certainly be an online rollback segment after switching, O (partition _ partition) O ha!
Try to use the _ smu_debug_mode method to make the rollback segment offline. As a result, the result fails as always. In the end, there is no way to use the most useful hidden parameters:
_ Offline_rollback_segments
_ Uploupted_rollback_segments
After the settings, restart the database and recreate the undo tablespace. The database returns to normal and no error 600 is reported.
A short answer record. Thank you very much, boss. She just left the site when she was about to finish this article. Due to the relationship between vpn duty, I have been at home, but the boss has been at the site.