Redo Space Wait Ratio

[pareshg]

Hello Guys,

I am checking server statistics in TOAD which is showing Redo Space Wait Ratio = 199.09. What does it mean exactly ?

Oracle 8.1.7 on Win NT.
I have 4 redolog files of 10MB each.

Regards

Paresh

[stancho]

How many times in an hour do you have redo log switch? If it less then 15 minutes you should make your redo logs bigger (in 9i by default they are 3 groups 100MB each, you can try with something simmilar).

[julian]

Originally posted by pareshg
Hello Guys,

I am checking server statistics in TOAD which is showing Redo Space Wait Ratio = 199.09. What does it mean exactly ?

Oracle 8.1.7 on Win NT.
I have 4 redolog files of 10MB each.

Regards

Paresh

4 redolog files? Is this 2 groups each with 2 members or 4 groups each with exactly one member?

The wait indicates the time spent on waiting on disk space to be allocated for redo.

Do as Stantcho suggested.

What values do you have for LOG_CHECKPOINT_INTERVAL and LOG_CHECKPOINT_TIMEOUT?

[stancho]

Originally posted by julian
What values do you have for LOG_CHECKPOINT_INTERVAL and LOG_CHECKPOINT_TIMEOUT?

Julian if those 2 parameters are incorrect I think the sessions will wait for checkpoint ?

And greetings from Bulgaria

[pareshg]

Julian,

Values are :
LOG_CHECKPOINT_INTERVAL=10000
and LOG_CHECKPOINT_TIMEOUT=1800

Also I have 4 groups with 1 member each. Log switching occurs almost after 15 mins. But its occur i think because of Materialized Views. There are no of MV which does complete refresh from other DB server after 15 mins. There are no high no users accessing database.

Is there any way to increase timing between archived files and thus reducing no of archives files generated.

Thanks

Paresh

[julian]

Originally posted by stancho
Julian if those 2 parameters are incorrect I think the sessions will wait for checkpoint ?

Yes, checkpoints will happen often and this will cause file header syncronization to take place often too. It may be a matter of moving his/her online logs to another disk. It could be that the current disks are too busy. This should be avoided.

And greetings from Bulgaria

Thanks :-)