Hi folks,
I have suffered a nightmare this last night when my production database hangs during a hot backup. There are the traces on the log:
ALTER TABLESPACE SYSTEM BEGIN BACKUP
Thu Oct 25 21:00:03 2007
Completed: ALTER TABLESPACE SYSTEM BEGIN BACKUP
Thu Oct 25 22:06:45 2007
WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=47
and on the dump file generated:
*** SESSION ID:(45.93) 2007-10-25 22:06:45.036
WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! <<<
row cache enqueue: session: 3F296138, mode: N, request: S
row cache parent object: address=40F3A440 cid=3(dc_rollback_segments)
hash=3664ff5a typ=5 transaction=3FC7416C flags=00000002
own=40F3A4B4[3FEE5AA4,3FEE5AA4] wat=40F3A4BC[3FEE5A14,3FEE5B04] mode=X
status=VALID/-/-/-/-/-/-/-/-
data=
0000000b 0000000b 0000000b 00000009 535f000a 4d535359 24313155 00000000
00000000 00000000 00000000 00000000 00000001 00000001 00000001 00000002
0002dd6f 00000000 00000000 00000001
owners:
Fortunately, I have been able to restart the database after restablish the tablespace mode to NORMAL. What I don't understand why that has happened. The hot backup has been working for a long time without troubles.
The dump files signals something around the rollback_segment but my database works is running in automatic undo mode. What I'm doing wrong?
Thanks by advance,
Alfons