Hi guys
I am benchmarking my dell r610 with 6x300gb 2.5 10k Drives in raid10 using IOMETER.
I am using the H700 controller 512 cache.
Using a 100% read test.
I am getting a 22.6K IOPS
Which is pretty much the same number I get with 4x300GB drives in raid10. Am I saturating the controller? I was thinking I would get 30K IOPS.
TIA
Doesn't seem to far off IMO.
http://lonesysadmin.net/2010/03/30/dell-poweredge-r610-perc6i-disk-comparison/
Apparently Raid 5 is really smoking... which makes sense...
You would achieve even better performace if you did use the 15k Drives but due to the larger caching on the bigger drives may be whats helping in that area.
br>
Cheers,
Chad King
VCP-410 | Server+
Twitter: http://twitter.com/cwjking
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
RAId 5 would give you much better read performance, but honestly, the READ performance is not where people usually run into trouble, its on the writes, especially if the controller doesn't have a BBWC on it (although I think the H700 does).
If you need performance, then RAID10 will probably give you the best overall balance of read/write, but if you need capacity, then RAID 5 wll so, but you will have lower write scores (but better read then RAID 10)