Exchange 2003 I/O performance issues

Microsoft Exchange 2003
Hi, Reading Microsoft recommendation, the latency (sec/read or sec/write) for the Exchange database disk (separated from the transaction log disk) should not go above 40 ms average and 100 ms for spike. On my Exchange 2003 server I?m getting 122 ms for the average and 257 ms for the spike. The disk subsystem in my environment includes: two FC HBA?s, each connected using FC cables to a controller ( one HBA to one controller, total of two sets) on an FC disk array (Raidsys). On the FC disk array I have two LUN each with RAID 5 and each LUN expose via one controller ( total of two LUN). From the server side the OS seas two disks, one for the database and one for the transaction log. Am I?m doing something wrong? Thanks!

Answer Wiki

Thanks. We'll let you know when a new response is added.

How many drives in your raid5 array where the IO is performing badly? In my experience, raid5 is not the best array type for this activity. I would suggest a test setup using raid 0+1 (striping + mirroring) for best IO performance results.

Discuss This Question: 1  Reply

There was an error processing your information. Please try again later.
Thanks. We'll let you know when a new response is added.
Send me notifications when members answer or reply to this question.
  • Dcollimore
    When sizing disk layouts for Exchange, spindle count is VERY important. You say you have 2 disks, 1 for the DB and 1 for the transaction logs. However, you do not say what kind of drive. 15K 146Gb SCSI drives are pushing around 150 to 200 IOPS in a RAID 1+0 configuration, and that number drops to about 110-130 in a RAID 5 scenario. 10K drives oush even less IOPS, so it is important to know exactly what your disks are giving you. If you check your perfmon output, you will see what your disks are pushing and what is being require by Exchange. Sounds to me as if you do not have enough spindles behind your database. Are you also seeing SMTP queueing issues? That should also have its own disk. Also, there are two papers by Microsoft that get into details of what I have described including what counters needs to be monitored to obtain these #s. The first one is entitled-Exchange Server 2003 Performance and Scalability. The second, which from your email I believe you have read, is entitled-Optimizing Storage for Exchange 2003. Douglas Collimore
    0 pointsBadges:

Forgot Password

No problem! Submit your e-mail address below. We'll send you an e-mail containing your password.

Your password has been sent to:

To follow this tag...

There was an error processing your information. Please try again later.

Thanks! We'll email you when relevant content is added and updated.


Share this item with your network: