in the last 4 days we have the strange situation that the autolog functionality is not working and the log devspace gets filled up.
"Normal" behaviour is, that every 10min a log backup is being created and copied to a second server. But in the last 4 days we have the situation (once a day) that this doesn't work and the log-devspace gets filled up.
Even a log backup doesn't solve the situation.
The only "solution" is restarting the database, which isn't a nice "solution" for a productive system.....
Today we restarted the database at 02:13pm. In the dbm.-files there are no error messages regarding the 10min-interval-creation of autologs, but we had to restart the db since the log devspace got filled up to 80% ?!?!
The dbm.utl/dbm.prt files show entries every 10min. for creating the autolog file, but in reality it didn't have been written..
What else can we check to get closer to the source of this problem ?
I can only think of a long running transaction which can cause such a situation, but we have no long running....?!?!
What exactly can we check if we're running in such a situation again ?
many thanks in advance...GERD...
2009-11-23 13:56:51 0x00002d59 0 DBM command autolog_on 2009-11-23 14:06:56 0x00007bff 0 DBM command autolog_off 2009-11-23 14:07:01 0x000001d7 0 DBM command backup_start auto 2009-11-23 14:07:02 0x00000260 0 DBM command autolog_on 2009-11-23 14:08:03 0x00007a82 0 DBM command db_offline 2009-11-23 14:13:07 0x00007a82 0 DBM command db_online 2009-11-23 14:17:59 0x00001f20 0 DBM command autolog_off 2009-11-23 14:18:04 0x0000243a 0 DBM command backup_start auto 2009-11-23 14:18:05 0x00002521 0 DBM command autolog_on 2009-11-23 14:28:10 0x00004593 0 DBM command autolog_off 2009-11-23 14:28:15 0x0000498b 0 DBM command backup_start auto 2009-11-23 14:28:16 0x00004a21 0 DBM command autolog_on 2009-11-23 14:38:27 0x000074a5 0 DBM command autolog_off 2009-11-23 14:38:32 0x00007801 0 DBM command backup_start auto 2009-11-23 14:38:33 0x0000788a 0 DBM command autolog_on