-
Our Unix admins (HP-UX-11) will be relaying the disks used for our datawarehouse system (Oracle 8.1.7 ) that is currently using 8K block size.
I am wondering if I should suggest them the stripe size that will be good for my database performance (I am not sure yet if I can do that ); on the other hand I wonder that if I already know that their stripe size will be a multiple of 8K then all I need to do is leave the block size alone (because Export/Import may take long time in order to change the block size) and set the DB_FILE_MULTIBLOCK_READ_COUNT parameter such that DB_BLOCK_SIZE * DB_FILE_MULTIBLOCK_READ_COUNT equals the stripe size.
Is there any difference in (performance wise) setting the block_size equal to the stripe size or setting the block_size * multiblock_read_count equal to the stripe size?
Thanks,
- Rajeev
Rajeev Suri
-
Your db_block_size should be a multiple of your OS block size. If you have 2K OS blocks, an 8K db_block_size should be fine.
Your stripe size should be at least equal to your maximum Oracle I/O size. Your maximum I/O is db_block_size * db_file_multiblock_read_count.
See http://metalink.oracle.com/metalink/... ITS EFFECT ON for details.
Jeff Hunter
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|