Okay, HDS has updated the AMS line with some new arrays as per this article and utilising SAS disks; not a bad idea and they look okay, not earth-shattering; but hey this is HDS and I don’t really expect anything earth-shattering. But this comment from the article made my eyebrows raise;
HDS has rounded up an IDC analyst, Richard Villars, who says that the combination of SAS backplane and better controller pairing makes the new AMS products well suited to supporting virtualised servers. This is because they can be provisioned faster, have better performance, etc. etc
How does a SAS backplane and better controller pairing mean that that they can be provisioned faster? Now better software might but surely not a change in backplane? And how this all leads to being able to support virtualised servers better? I’m really not sure!! What about you?
Very odd comment, and Chris Mellor is normally quite sharp with this; perhaps a touch of sarcasm in the etc etc?
Why don’t you ask Chris on the Register directly? He emailed me back pretty promptly the last tme I left a comment.
On a general note, any array without the minimum of thin cloning makes it a non-starter for any serious server virtualisation. HDS aren’t clear on this; will their thin provisioing include it?
Do you mean thin provisioning or chubby provisioning??
LOL! That was 3Par’s Crag Nunes about HP’s so called thin provisioning. It will be interesting to see what variety it turns out to be.
Martin — not surprisingly, I was pretty acerbic about this announcement on my blog.
Interestingly enough, I was able to coax Hu out of his blogging monologue for a direct (albeit loopy) response.
My bottom line? They should keep the focus on the high end, where they at least have a position.