« Fact or fiction? CHP quotas and thin provisioning | Main | Calculating the output of storage tiering in wide striping arrays »

March 31, 2010

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

nate

you using that iops calculator spreadsheet to come up with those #s ? Since I think it defaults to 250 IOPS for 15k RPM, and 90 for SATA.

Is your RAID 5 there 3+1 ? I find the RAID6 numbers interesting, RAID6 isn't available in the IOPS spreadsheet I have just confirms I made a good choice by not going RAID 6 :) [running mostly RAID 5 2+1 with some 5+1]

XIV looks interesting but I lost interest fast when I saw it tops out at 180 disks, in this day & age 180 disks is not enough, even if you don't need it right away, having the system at least support say ~400+ disks is a nice cushion to have if your starting out on the smaller side.

marc farley

Hi Nate,

I don't think I used the same IOPS spreadsheet and I plugged lower values into the one I had simply because I wanted to post conservative estimates that people would easily agree with. Also, I'm fairly certain the RAID 5 is modeled after 3+1 Raid sets, but it was not spelled out. So this was not an attempt to be very precise - I was looking to publish good ballpark numbers that would help readers understand the larger effect of wide striping.

Yeah, the RAID 6 numbers are interesting. FWIW, I'm working on another post that shows the IOPS potential for arrays with storage tiering enabled.

The comments to this entry are closed.

Search StorageRap


Subscribe

Latest tweets

3PARTV

  • Loading...

Blogroll

Infosmack Podcasts

Virtumania Podcasts

Subscribe