{"created":"2025-01-19T01:12:25.819478+00:00","updated":"2025-01-19T17:52:00.326235+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00211245","sets":["1164:1867:10528:10599"]},"path":["10599"],"owner":"44499","recid":"211245","title":["仮想NUMAマシンの性能及び弾力性の向上"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-05-20"},"_buckets":{"deposit":"10317290-708e-452b-a894-c416c03d0c74"},"_deposit":{"id":"211245","pid":{"type":"depid","value":"211245","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"仮想NUMAマシンの性能及び弾力性の向上","author_link":["536340","536341","536342"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"仮想NUMAマシンの性能及び弾力性の向上"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"性能改善","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-05-20","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"東京大学"},{"subitem_text_value":"東京大学"},{"subitem_text_value":"東京大学"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"The University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"The University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"The University of Tokyo","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/211245/files/IPSJ-OS21152010.pdf","label":"IPSJ-OS21152010.pdf"},"date":[{"dateType":"Available","dateValue":"2023-05-20"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-OS21152010.pdf","filesize":[{"value":"1.2 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"11"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"135f04fa-e02b-4967-8456-330c0bdfeb09","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"味曽野, 雅史"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"林, 遼"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"品川, 高廣"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10444176","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8795","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"NUMA マシンの普及に伴い，仮想化環境でも NUMA マシンを利用することが一般的になってきている．NUMA マシン上で仮想化をおこなう一つの方法が，ゲストにホストの NUMA 構成を一部または完全に再現する仮想 NUMA (vNUMA; Virtual NUMA) である．このときゲスト上では実マシンの NUMA 構成を活用したスケジューリングやメモリ利用が可能となる．vNUMA で VM を作成する場合，ホストの NUMA トポロジを完全にゲスト上で再現した場合（Full vNUMA と呼ぶ）が最も性能が良くなると考えられるが，Full vNUMA に関して詳細に性能を調査した研究は少ない．また，Full vNUMA 環境で他の VM を作成した場合は必然的にオーバコミットが発生するため，vNUMA の優位性が低くなるという課題がある．本論文ではまず Linux の QEMU/KVM を利用して Full vNUMA 環境を作成し，その性能評価をおこなった．この結果．我々は現在の Linux には，vNUMA 環境において，1) idle 時の halt 状態時の不適切な PV 機能の処理による vCPU の利用率の低下，2) idle 時の過渡なプロセスマイグレーションによる性能低下の問題が生じることを発見した．この問題の解決策として，PV 機能の修正及び idle 時のマイグレーションを抑制する方法を提案する．また，Full vNUMA 構成で複数 VM を作成した場合のオーバコミットを避けるための方法として，VM 間で協調的かつ排他的にリソースを共有する方法を提案する．これにより Full vNUMA の利点を生かしつつ，状況に応じて柔軟に複数 VM によりリソースを活用することができるようになる．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"12","bibliographic_titles":[{"bibliographic_title":"研究報告システムソフトウェアとオペレーティング・システム（OS）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-05-20","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"10","bibliographicVolumeNumber":"2021-OS-152"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":211245,"links":{}}