Progress Database layout

[Nick Lewers]

HPUX 11i 64bit
VA7410

I have a database that is around 120gb. I currently have around 300gb available in my VA, 150gb in each redundancy group. I would like to ensure that IO is spread evenly across all disks in the array. So I was thinkging about creating a 75gb lun on each RG and creating one large filesystem striping across both luns.

Is there any downside to doing that that I should be aware of? Is it recommended to do this or is it perhaps better to physically layout the database files to each of the luns?

[Steven E. Protter]

Shalom,

The downside is your dba's still need to cooperate with you to make this layout work.

redo logs work best with either raid 1 storage or raid 5 spread over a lot of disks. I don't see how many disks you have on the VA7410, but my experience is these are not very fast disks.

archive logs are sequential and can work very well on raid 5. The need to spread the i/o over a lot of disks is less.

I'd suggest going to Oracle's technet website and looking into the ways to get maximum disk performance.

SEP

[Keith Jahn]

Keep the Progress AI and BI files on distinct luns and try to spread the database segments evenly across the two luns.