Printable View
Hi all,
I turned on archive logging on to see if it works correctly but I noticed that every minutes i get 20 megs of archive file and I am going crazy trying to manage them. See every night at 12:00am I run hot backups on the box and I expect to clear up the mountpoint /arch1 4gig with a cron job but I need some guidance in tuining the archive logging. I dont want 20 megs a minute.
pls advice asap
Amount of archive logs depends on your database activity and a couple of parameters in init.ora Check for log_checkpoint_interval and log_checkpoint_timeout in init.ora
F.e. on 1.6Tb database with high user activity I have up to 8 archive logs (10Mb each) per minute.
Good luck!
Sergey.
The thumb rule is to size the redolog files such that there would be 3-4 log switches per hour.
Hi,
Check the values of Log_Checkpoint_Interval and Log_Checkpoint_Timeout and set them appropriately. If still the problem persists, then you may have to check your transactions.
Regards
Ganesh .R
log_checkpoint_interval
41000
log_checkpoint_timeout
0
log_checkpoints_to_alert
FALSE
Hello. I have no redo in the morning because its a hybrid database.
Please suggest new values for me.
thanks
i have 4 groups with 20meg 2 members.
Jigar
you can't do much about the volume of redo info you generate. you'll generate whatever is needed by the users. the values of Log_Checkpoint_Interval and Log_Checkpoint_Timeout will not affect the amount generated nor how many redo logs you use. these parameters only affect checkpoints, not logswitches. you can limit some operations' redo information with NOLOGGING (or UNRECOVERABLE) but concern about redo generation is a REALLY bad reason to use these clauses.
whether or not you run in archivelog mode, you should not turn over 60 redo logs/hour. increase the size of your redo logs (look at ALTER DATABASE ADD LOGFILE and then ALTER DATABASE DROP LOGFILE; use in conjunction with V$LOG and V$LOGFILE). be careful when managing redolog files in the operating system.
d.
check the size of your REDO logs - if they're too small, they fill up fast - each time they fill up a log switch is required. Each log switch generates an archive file.
Dear jlakhani, 15th April 2001 15:11 hrs chennai
I would like to know the DB size and nature of DB activity etc.
As DBAtrix has said there is much can be done with the prameters you have provided.
The only thing you can do is using NOLOGGING options for some transactions.
Cheers
Padmam
hello,.....padmam...db size is about 30 gig. its an oltp system.
I currently have 4 groups with 20 megs / member.
shall i increase the memberst o 50 megs? Why whould I use nologging. BEcause i needt to be able to recover if my db crashes thats why I have archiving....