{"id":165161,"date":"2020-02-28T19:01:22","date_gmt":"2012-04-01T00:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/msr-research-item\/pacman-coordinated-memory-caching-for-parallel-jobs\/"},"modified":"2021-12-07T14:51:45","modified_gmt":"2021-12-07T22:51:45","slug":"pacman-coordinated-memory-caching-for-parallel-jobs","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/pacman-coordinated-memory-caching-for-parallel-jobs\/","title":{"rendered":"PACMan: Coordinated Memory Caching for Parallel Jobs"},"content":{"rendered":"<div class=\"asset-content\">\n<p>Data-intensive analytics on large clusters is important for modern Internet services. As machines in these clusters have large memories, in-memory caching of inputs is an e\u009affective way to speed up these analytics jobs. The key challenge, however, is that these jobs run multiple tasks in parallel and a job is sped up only when inputs of all such parallel tasks are cached. Indeed, a single task whose input is not cached can slow down the entire job. To meet this \u201call-or-nothing\u201d property, we have built PACMan, a caching service that coordinates access to the distributed caches. is coordination is essential to improve job completion times and cluster effi\u009bciency. To this end, we have implemented two cache replacement policies on top of PACMan\u2019s coordinated infrastructure\u2013LIFE that minimizes average completion time by evicting large incomplete inputs, and LFU-F that maximizes cluster effi\u009bciency by evicting less frequently accessed inputs. Evaluations on production workloads from Facebook and\u009f Bing show that PACMan reduces average completion time of jobs by 53% and 51% (small interactive jobs improve by 77%), and improves e\u009bfficiency of the cluster by 47% and 54%, respectively.<\/p>\n<\/div>\n<p><!-- .asset-content --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Data-intensive analytics on large clusters is important for modern Internet services. As machines in these clusters have large memories, in-memory caching of inputs is an e\u009affective way to speed up these analytics jobs. The key challenge, however, is that these jobs run multiple tasks in parallel and a job is sped up only when inputs [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13563,13547],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-165161","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-data-platform-analytics","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2012-4-1","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_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":"206060","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/pacman_nsdi12.pdf","id":"206060","title":"pacman_nsdi12.pdf","label_id":"243109","label":0}],"msr_related_uploader":"","msr_attachments":[{"id":206060,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/02\/pacman_nsdi12.pdf"}],"msr-author-ordering":[{"type":"user_nicename","value":"Ganesh Ananthanarayanan","user_id":31834,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Ganesh Ananthanarayanan"},{"type":"text","value":"Ali Ghodsi","user_id":0,"rest_url":false},{"type":"text","value":"Andrew Wang","user_id":0,"rest_url":false},{"type":"text","value":"Dhruba Borthakur","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Srikanth Kandula","user_id":33707,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Srikanth Kandula"},{"type":"text","value":"Scott Shenker","user_id":0,"rest_url":false},{"type":"text","value":"Ion Stoica","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[199565],"msr_event":[],"msr_group":[144899],"msr_project":[239762],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/165161"}],"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\/165161\/revisions"}],"predecessor-version":[{"id":537941,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/165161\/revisions\/537941"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=165161"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=165161"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=165161"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=165161"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=165161"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=165161"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=165161"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=165161"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=165161"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=165161"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=165161"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=165161"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=165161"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=165161"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=165161"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}