{"id":9298,"date":"2008-12-09T08:57:00","date_gmt":"1999-11-29T20:00:00","guid":{"rendered":"https:\/\/aidanfinn.com\/?p=9298"},"modified":"2008-12-09T08:57:00","modified_gmt":"1999-11-29T20:00:00","slug":"distributed-power-management","status":"publish","type":"post","link":"https:\/\/aidanfinn.com\/?p=9298","title":{"rendered":"Distributed Power Management"},"content":{"rendered":"<p>Most administrators don&#8217;t know or care about the <em>real<\/em> cost of servers: power.\u00a0 A single server&#8217;s cost is <em>much<\/em> more than what you pay to Dell or HP.\u00a0 The power alone massively outweighs the purchase cost.\u00a0 It&#8217;s said a typical server has the carbon footprint of a car.\u00a0 It&#8217;s inevitable that we&#8217;re going to see carbon taxes hitting businesses now.\u00a0 Cloud computing\/Software-as-a-Service mightn&#8217;t be for everyone so they need a solution.\u00a0 Cloud providers also need a solution to power issues because the biggest cost they have to pass on to customers is electricity.<\/p>\n<p>I found this <a title=\"commentary \" href=\"http:\/\/dcsblog.burtongroup.com\/data_center_strategies\/2008\/11\/distributed-power-management-improved-efficiency-or-playing-with-fire.html\">commentary <\/a>by Chris Wolf talking about an experimental feature that was included in VMware VI3.5.\u00a0 This feature called Distributed Power Management (DPM) is an interesting one &#8211; one which had me nearly swinging towards VMware instead of Hyper-V.\u00a0 Virtual Center monitors the usage of host resources by VM&#8217;s and using DRA and memory over-subscription it will consolidate VM&#8217;s to fewer hosts.\u00a0 This allows idle hosts to be powered down or suspended.\u00a0 When resource consumption grows the required idle hosts are powered back up using WOL.\u00a0 VM&#8217;s can be migrated using VMotion to ensure they get the CPU and RAM (probably IO as well) resources that they need.<\/p>\n<p>The commentary talks about how people are wary of powering down\/up production servers.\u00a0 That&#8217;s fair enough.\u00a0 In my opinion however, that&#8217;s the wrong way to look at this.\u00a0 The production servers are the VM&#8217;s.\u00a0 In this scenario the VM&#8217;s are never powered down.\u00a0 They&#8217;re offline for a few milliseconds as the VMotion across the cluster, something that VMware customers are well used to now.\u00a0 <\/p>\n<p>The hosts are just physical resources.\u00a0 The hardware is just an enabling layer like electricity or network when you&#8217;re dealing with virtualisation.\u00a0 And just like those utilities there&#8217;s fault tolerance at this layer &#8211; or there should be.\u00a0 In a network that could realistically use DPM to save power there will be significant numbers of hosts.\u00a0 They should be dealing with <em>at least<\/em> N+1 the number of hosts that they require, maybe even N+2.\u00a0 So what happens if there&#8217;s an occasional hardware failure?\u00a0 If you run an enterprise network then the hardware should be monitored and any faults will be responded to immediately.<\/p>\n<p>Microsoft are currently taking a different approach to the power issue when it comes to Windows Server 2008 R2 &#8211; and logically Hyper-V.\u00a0 MS are using Core and CPU Parking.\u00a0 The server monitors the demand on the CPU cores every X milliseconds.\u00a0 When a core is idle it is suspended, thus reducing it&#8217;s power consumption.\u00a0\u00a0 The CPU core is the major draw on power in a server.\u00a0 It&#8217;s also the generator of heat and cooling that heat is another major draw on power.\u00a0 Suspending idle Cores reduces both of those power demands.\u00a0 If a Core is required then it is snapped back online.\u00a0 The trick is in defining appropriate idle windows &#8211; you don&#8217;t want to suspend at millisecond 1 and find you&#8217;re always bringing it back online at millisecond 2.\u00a0 That&#8217;s wasteful.\u00a0 When all cores in a CPU are idle then the CPU is parked, thus saving more power.<\/p>\n<p>I was at a power meeting\/interview session with MS at TechEd EMEA and I brought up the VMware DPM approach.\u00a0 I don&#8217;t know if it&#8217;s something MS will look at or not.\u00a0 I hope they do look at it for the next release after Windows Server 2008 R2.\u00a0 Right now, I have to applaud VMware for trying to do something.\u00a0 They do see the hardware as just an enabling layer, not the production servers.\u00a0 I think that&#8217;s the right point of view to take.\u00a0 When DPM does go live I can see it saving VMware customers a good bit of money.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Most administrators don&#8217;t know or care about the real cost of servers: power.\u00a0 A single server&#8217;s cost is much more than what you pay to Dell or HP.\u00a0 The power alone massively outweighs the purchase cost.\u00a0 It&#8217;s said a typical server has the carbon footprint of a car.\u00a0 It&#8217;s inevitable that we&#8217;re going to see &hellip; <a href=\"https:\/\/aidanfinn.com\/?p=9298\" class=\"more-link\">Continue reading<span class=\"screen-reader-text\"> &#8220;Distributed Power Management&#8221;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"_uf_show_specific_survey":0,"_uf_disable_surveys":false,"footnotes":""},"categories":[20],"tags":[],"class_list":["post-9298","post","type-post","status-publish","format-standard","hentry","category-hyper-v"],"aioseo_notices":[],"jetpack_featured_media_url":"","amp_enabled":true,"_links":{"self":[{"href":"https:\/\/aidanfinn.com\/index.php?rest_route=\/wp\/v2\/posts\/9298","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/aidanfinn.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/aidanfinn.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/aidanfinn.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/aidanfinn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=9298"}],"version-history":[{"count":0,"href":"https:\/\/aidanfinn.com\/index.php?rest_route=\/wp\/v2\/posts\/9298\/revisions"}],"wp:attachment":[{"href":"https:\/\/aidanfinn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=9298"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/aidanfinn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=9298"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/aidanfinn.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=9298"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}