This question goes out to the Digital Unix (4.0D) / Oracle 7.3.3 users:
We are ordering hardware to run as our mission critical UNIX/Oracle Server.
An AlphaServer 1200 Dual CPU(SMP) with 256MB RAM has been chosen.
This should provide more than enough throughput for our single instance Oracle DB.
Our database actively uses a couple of tablespaces of less than 4GB total. This data
is moved off monthly to a separate tablespace and very rarely accessed (read only necessary).
A years worth of data is kept, and then deleted. The past 11 month's worth of data is
calculated to be at approximately 50GB. We plan to get 90GB. So.. here's the question(s):
How can we organize our disks so that we can optimize disk access and reliability?
Optimally, the tablespaces, DU, log files, and swap are each on a separate I/O channel/controller.
Optimally the tablespaces are on a raw disk and Oracle can perform asynchronous I/O (AIO).
Should we use RAID 5 for reliability and three RAID controllers to keep the data flow efficient?
Can we use the raw disk/Async I/O function with the HW RAID controllers?
Is Ultra SCSI the way to go?
Comments would be greatly appeciated.
Thanks,
Wayne
***********************
Wayne Sweatt
Principal Software Analyst
Litton / PRC
505.827.9288
***********************
Received on Thu Dec 31 1998 - 18:18:15 NZDT