LOG_CHECKPOINT_INTERVAL

[kgb]

Hi guys!
I have a question:
I have 2 partition on linux box: first one has block size 1kb, second one 4kb. Oracle and OS located in first partition, DB files in second partition. What the block size for LOG_CHECKPOINT_INTERVAL parameter?
Thanks!

[dknight]

Database checkpoints happen whenever the value of the LOG_CHECKPOINT_INTERVAL is exceeded.

The value is the number of blocks. So if you want a checkpoint every 8m bytes and your blocksize is 8k, the value would be:

8m/8k = 1k blocks

Good luck.

[kgb]

LOG_CHECKPOINT_INTERVAL is in OS blocks, so what the OS block size for this value if I have 1kb for first partition and 4 Kb for second?

Originally posted by dknight
Database checkpoints happen whenever the value of the LOG_CHECKPOINT_INTERVAL is exceeded.
The value is the number of blocks. So if you want a checkpoint every 8m bytes and your blocksize is 8k, the value would be:
8m/8k = 1k blocks
Good luck.

[dknight]

It is the number of blocks of redo generated before a checkpoint occurs.

OS block is independent of the partitions you create.

[dknight]

This is clear. Please ignore my last response. I do not know the answer.

Depending on where your redo logs are stored, you can find the answer with trial and error.

Good luck.

[kgb]

LOG_CHECKPOINT_INTERVAL - this is set to the number of redo log blocks written between checkpoints. The blocks used in sizing this parameter are operating system blocks, NOT Oracle buffer blocks.

Originally posted by dknight
It is the number of blocks of redo generated before a checkpoint occurs.
OS block is independent of the partitions you create.

[dknight]

kgb,

My last response is incorrect. You are absolutely right about OS blocks.

Good luck.