Developers use Machine Learning (ML) platforms to train ML models and then deploy these ML models as web services for inference (prediction). A key challenge for platform providers is to guarantee response-time Service Level Agreements (SLAs) for inference workloads while maximizing resource efficiency. Swayam is a fully distributed autoscaling framework that exploits characteristics of production ML inference workloads to deliver on the dual challenge of resource efficiency and SLA compliance. Our key contributions are (1) model-based autoscaling that takes into account SLAs and ML inference workload characteristics, (2) a distributed protocol that uses partial load information and prediction at frontends to provision new service instances, and (3) a backend self-decommissioning protocol for service instances. We evaluate Swayam on 15 popular services that were hosted on a production ML-as-a-service platform, for the following service-specific SLAs: for each service, at least 99% of requests must complete within the response-time threshold. Compared to a clairvoyant autoscaler that always satisfies the SLAs (i.e., even if there is a burst in the request rates), Swayam decreases resource utilization by up to 27%, while meeting the service-specific SLAs over 96% of the time during a three-hour window. Microsoft Azure\u2019s Swayam-based framework was deployed in 2016 and has hosted over 100,000 services.<\/p>\n","protected":false},"excerpt":{"rendered":"
Developers use Machine Learning (ML) platforms to train ML models and then deploy these ML models as web services for inference (prediction). A key challenge for platform providers is to guarantee response-time Service Level Agreements (SLAs) for inference workloads while maximizing resource efficiency. Swayam is a fully distributed autoscaling framework that exploits characteristics of production […]<\/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":[246574],"research-area":[13547],"msr-publication-type":[193716],"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":[264846],"msr-pillar":[],"class_list":["post-476721","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-highlight-award","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2017-12-11","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":"Best Student Paper Award, Middleware\u201917","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":"454515","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2018\/01\/2017.Middleware.Swayam.TailLatencyInAzureML.pdf","id":"454515","title":"2017.Middleware.Swayam.TailLatencyInAzureML","label_id":"243109","label":0},{"type":"doi","viewUrl":"false","id":"false","title":"10.1145\/3135974.3135993","label_id":"243109","label":0}],"msr_related_uploader":"","msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Arpan Gujarati","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Sameh Elnikety","user_id":33503,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Sameh Elnikety"},{"type":"user_nicename","value":"Yuxiong He","user_id":35084,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Yuxiong He"},{"type":"text","value":"Kathryn S. McKinley","user_id":0,"rest_url":false},{"type":"text","value":"Bjorn B. Brandenburg","user_id":0,"rest_url":false}],"msr_impact_theme":["Computing foundations"],"msr_research_lab":[199565],"msr_event":[],"msr_group":[144927],"msr_project":[],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/476721"}],"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":3,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/476721\/revisions"}],"predecessor-version":[{"id":1028862,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/476721\/revisions\/1028862"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=476721"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=476721"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=476721"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=476721"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=476721"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=476721"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=476721"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=476721"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=476721"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=476721"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=476721"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=476721"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=476721"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=476721"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=476721"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=476721"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}