Comments on: Should I buy blades or standalones for my virtualization platform? http://vmguy.com/wordpress/index.php/archives/66 Virtualization for the little guy Fri, 27 Jan 2012 15:54:17 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Dave Lawrence http://vmguy.com/wordpress/index.php/archives/66/comment-page-1#comment-9 Dave Lawrence Fri, 14 Nov 2008 13:49:17 +0000 http://VMGUY.COM/wordpress/?p=66#comment-9 Massimo, great point. I intended to write chassis (no more blogging after midnight.) I would agree that the Ethernet Switches and/or FC Switches on the back of the chassis are an unfair comparison. You are absolutely correct in that these switches would not be turned off by DPM and thus are not a true "apples to apples" comparison. In my previous tenure as an administrator, I only used HP blades. Unfortunately, HP's Blade Power Calculator does not let you pull out the switches off the enclosure when computing power requirements (they do but it's not a real world option). I retract my statement about the power savings and I will write a separate upcoming comparison between the two, I'll use more of a real world scenario (including the power for the switches in both scenarios if I have to.) Thanks for keeping me honest. I've struck out the power section of this post until I can follow up with a better real-world apples to apples comparison. Massimo, great point. I intended to write chassis (no more blogging after midnight.) I would agree that the Ethernet Switches and/or FC Switches on the back of the chassis are an unfair comparison. You are absolutely correct in that these switches would not be turned off by DPM and thus are not a true “apples to apples” comparison. In my previous tenure as an administrator, I only used HP blades. Unfortunately, HP’s Blade Power Calculator does not let you pull out the switches off the enclosure when computing power requirements (they do but it’s not a real world option). I retract my statement about the power savings and I will write a separate upcoming comparison between the two, I’ll use more of a real world scenario (including the power for the switches in both scenarios if I have to.) Thanks for keeping me honest. I’ve struck out the power section of this post until I can follow up with a better real-world apples to apples comparison.

]]> By: Massimo Re Ferre' http://vmguy.com/wordpress/index.php/archives/66/comment-page-1#comment-7 Massimo Re Ferre' Fri, 14 Nov 2008 08:34:11 +0000 http://VMGUY.COM/wordpress/?p=66#comment-7 >I cannot reduce my power consumption as efficiently because so much of >the power is in the chassis backplane What do you exactly mean? The "backplane" doesn't consume any power. You might be referring to the switches in the back but if you are not using a chassis all of those 14/16 ports (depending which chassis you use IBM/HP) x 2 (for redundancy reasons) have to be available on your top of the rack switches .... and if DPM turns off your standalone servers those switches are not turned off so the "only 40% savings" is a weak argument. On a similar note you might be interested in this. This was my attempt: http://www.it20.info/misc/virtualizationplatformofchoice.htm Massimo. >I cannot reduce my power consumption as efficiently because so much of >the power is in the chassis backplane

What do you exactly mean? The “backplane” doesn’t consume any power. You might be referring to the switches in the back but if you are not using a chassis all of those 14/16 ports (depending which chassis you use IBM/HP) x 2 (for redundancy reasons) have to be available on your top of the rack switches …. and if DPM turns off your standalone servers those switches are not turned off so the “only 40% savings” is a weak argument.

On a similar note you might be interested in this. This was my attempt:
http://www.it20.info/misc/virtualizationplatformofchoice.htm

Massimo.

]]>