So a lot of people have asked for this. I cleaned up the doc a bit and fiddled the numbers to what I normally use.
http://lopoetve.com/RAID_Calc_Spreadsheet.xls
Have at it.
Take whatever drive you have, and google up the deep queue IOPS performance - I put in the numbers I use as a general benchmark, with the two extremes (almost any SATA disk can get ~100 IOPS now, and good SAS disks get ~280). Read/write are generic, change to match what you're doing based on application.
To work backwards, use the formulas in the lower right.
Remember, this is ~theoretical peak performance~. This does not include array overhead, ESX overhead / operations, or anything else. This is the absolute ~best~ case scenario - you will never actually reach this performance, but it will give you the ceiling.
Good luck
Let me know if you have questions.
http://lopoetve.com/RAID_Calc_Spreadsheet.xls
Have at it.
Take whatever drive you have, and google up the deep queue IOPS performance - I put in the numbers I use as a general benchmark, with the two extremes (almost any SATA disk can get ~100 IOPS now, and good SAS disks get ~280). Read/write are generic, change to match what you're doing based on application.
To work backwards, use the formulas in the lower right.
Remember, this is ~theoretical peak performance~. This does not include array overhead, ESX overhead / operations, or anything else. This is the absolute ~best~ case scenario - you will never actually reach this performance, but it will give you the ceiling.
Good luck