{"id":164107,"date":"2013-01-02T00:00:00","date_gmt":"2013-01-02T00:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/msr-research-item\/nobody-ever-got-fired-for-buying-a-cluster\/"},"modified":"2018-10-16T20:23:44","modified_gmt":"2018-10-17T03:23:44","slug":"nobody-ever-got-fired-for-buying-a-cluster","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/nobody-ever-got-fired-for-buying-a-cluster\/","title":{"rendered":"Nobody ever got fired for buying a cluster"},"content":{"rendered":"<div class=\"asset-content\">\n<p>In the last decade we have seen a huge deployment of cheap clusters to run data analytics workloads. The conventional wisdom in industry and academia is that scaling out using a cluster is better for these workloads than scaling up by adding more resources to a single server. Popular analytics infrastructures such as Hadoop are aimed at such a cluster scale-out environment, and in today&#8217;s world nobody gets fired for adopting a cluster solution.<\/p>\n<p>Is this the right approach? Our measurements as well as other recent work shows that the majority of real-world analytic jobs process less than 100GB of input, but popular infrastructures such as Hadoop\/MapReduce were originally designed for petascale processing. We claim that a single &#8220;scale-up&#8221; server can process each of these jobs and do as well or better than a cluster in terms of performance, cost, power, and server density. Is it time to consider the &#8220;common case&#8221; for &#8220;big data&#8221; analytics to be the single-server rather than the cluster case? If so, this has implications for data center hardware as well as software architectures.<\/p>\n<p>Unfortunately widely used platforms such as Hadoop perform poorly in a scale-up configuration. We describe several modifications to the Hadoop runtime to address this problem. These changes are transparent, do not require any changes to application code, and do not compromise scale-out performance. However they do significantly improve Hadoop&#8217;s scale-up performance. We present a broad evaluation across 11 representative Hadoop jobs that shows scale-up to be competitive in all cases and significantly better in some cases, than scale-out. Our evaluation considers raw performance, as well as performance per dollar and per watt.<\/p>\n<\/div>\n<p><!-- .asset-content --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>In the last decade we have seen a huge deployment of cheap clusters to run data analytics workloads. The conventional wisdom in industry and academia is that scaling out using a cluster is better for these workloads than scaling up by adding more resources to a single server. Popular analytics infrastructures such as Hadoop are [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13547],"msr-publication-type":[193718],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-164107","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"Microsoft Technical Report","msr_edition":"","msr_affiliation":"","msr_published_date":"2013-01-02","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"MSR-TR-2013-2","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"205604","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"msrtr-2013-2.pdf","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/msrtr-2013-2.pdf","id":205604,"label_id":0}],"msr_related_uploader":"","msr_attachments":[{"id":205604,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/msrtr-2013-2.pdf"}],"msr-author-ordering":[{"type":"text","value":"Raja Appuswamy","user_id":0,"rest_url":false},{"type":"user_nicename","value":"chrisgk","user_id":31424,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=chrisgk"},{"type":"user_nicename","value":"dnarayan","user_id":31659,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=dnarayan"},{"type":"user_nicename","value":"ohodson","user_id":33147,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=ohodson"},{"type":"user_nicename","value":"antr","user_id":31061,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=antr"}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"download":[],"msr_publication_type":"techreport","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/164107","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/164107\/revisions"}],"predecessor-version":[{"id":527539,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/164107\/revisions\/527539"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=164107"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=164107"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=164107"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=164107"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=164107"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=164107"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=164107"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=164107"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=164107"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=164107"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=164107"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=164107"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=164107"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=164107"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=164107"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=164107"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}