Not having such great luck with defrag programs in the past, I have stayed away from this type of software. When I had tried them before, they tended to be more problematic than they where worth. For starters, they hogged CPU, thrashed your hard drives constantly, and in some cases even caused corruption. Taking the lead from the review that Adrew Edney did on usingwindowshomeserver.com, I decided that I would give it another shot. My primary reasoning was that I was running these big arrays and wanted something that would actually work on them and keep them running in top condition. The native windows defrag simply does not work on large RAID volumes. The experience was not 100% flawless, however I must admit I was impressed, and the end results have been extremely good.
The program allows you tremendous options in configuration as well as providing you with smart data on your drives (Need SP1 installed).
This is what my drives looked like when it was done optimizing the first time. I excluded several drives/partitions such as the 100 meg partition on the primary drive, as well as the a 2T drive I use for backup, as it seemed to be pointless to run defrag on a backup drive (My Opinion).
The main issue I had was related to the Raid Mirrors both on my data drive and my OS drive (icydock with 2x 250 gig laptop drives). It seems that PerfectDisk does not read drives that are mirrored correctly and sets up “SMARTPlacement” as the default optimization configuration for those drives. On the RAID 5, it reads the drives as RAID uses “Consolidate Free Space” as the default method. The issue is that SMARTPlacement does not seem to work correctly on a mirror so you have to manually change the method for those drives to “Defrag Only” or “Consolidate Free Space” to avoid the hang ups and other problems. The other thing that choked me up a bit is the that by default the program tries to optimize the page file and the hibernation files. Again, a complete waste of time which takes forever to perform with little to no value. Fortunately you can excluded any folders or files you want from the process. In the end, with a little bit of customizing, you are up and running with end results being very good.
In the beginning I was skeptical that it would benefit me when using RAID. To my surprise, after my first optimization of my RAID array, I got a 10% gain in performance using ATTO as my benchmark. Despite the fact that I had to do a bit of tweaking, this application surprised me and turned out to be very stable and very efficient. It never used above 18% of CPU performance during optimization and it did not interfere with reading or writing to the arrays. When I was using my server during optimization, it gave no perceivable difference in performance. I have been running it for 3 weeks now on a schedule, and it has done a great job maintaining my RAID arrays. So far the software has lived up to my expectations and I certainly am happy that it will keep my arrays performing at their optimum.