Dennis Wood
Senior Member
Tim, big ups for this last article. I just wish you'd written it about 3 months ago as it would have saved me about 3 months worth of the goofing around with workstation setups, RAID, and the like
Thanks for the reference in the article too. Netgear's recommendation on using 3 drives in RAID0 on the workstation is exactly why all of our video editing workstations now use a minimum of 3 drives in RAID0, with the main workstation using the Adaptec 3405 in a 4 drive RAID0. What's really amazing now is that gigabit ethernet is so cheap...and we need 3 drive RAID0 arrays to max it out! I'm convinced now to the point that my company, Cinevate, is seriously looking at a new business unit.
Your findings here are very reassuring as they provide a body of evidence that parallel (as they should) what we've been doing here.
Now my only suggestion in your testing is to duplicate your single test machine and add load testing to your NAS charts. These type of test results had the Qnap folks tweaking the SAMBA configurations on their boxes...and the impact was profound. If a NAS unit is being sold for business use then it's performance should reflect grace under load
Consider also sourcing a MAC workstation (gasp) for your testing. Speaking of MACs, my Vista workstation died a few days ago and in 40 minutes (which included a 125GB restore) I was back up running. The disaster recovery strategy worked the charm using a Storagecraft recovery boot DVD and the workstation image sitting on the NAS. All I had to do is map a drive in the recovery environment and click restore.
Another topic for further exploration is the drives in the NAS. We're seeing a large difference between our two units, one using the WD "green" (cheap) drives and the other using the enterprise level Seagate 7200 rpm drives (twice the price). They're not on the same brand of switch...another variable.

Your findings here are very reassuring as they provide a body of evidence that parallel (as they should) what we've been doing here.
Now my only suggestion in your testing is to duplicate your single test machine and add load testing to your NAS charts. These type of test results had the Qnap folks tweaking the SAMBA configurations on their boxes...and the impact was profound. If a NAS unit is being sold for business use then it's performance should reflect grace under load

Another topic for further exploration is the drives in the NAS. We're seeing a large difference between our two units, one using the WD "green" (cheap) drives and the other using the enterprise level Seagate 7200 rpm drives (twice the price). They're not on the same brand of switch...another variable.
Last edited: