{"id":788258,"date":"2021-10-25T22:24:52","date_gmt":"2021-10-26T05:24:52","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-item&#038;p=788258"},"modified":"2021-10-25T22:24:52","modified_gmt":"2021-10-26T05:24:52","slug":"crashmonkey-and-ace-systematically-testing-file-system-crash-consistency","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/crashmonkey-and-ace-systematically-testing-file-system-crash-consistency\/","title":{"rendered":"Crashmonkey and ACE: Systematically testing file-system crash consistency"},"content":{"rendered":"<p>We present C<span class=\"gs_fscp\">rash<\/span>M<span class=\"gs_fscp\">onkey<\/span>\u00a0and A<span class=\"gs_fscp\">ce<\/span>, a set of tools to systematically find crash-consistency bugs in Linux file systems. C<span class=\"gs_fscp\">rash<\/span>M<span class=\"gs_fscp\">onkey<\/span>\u00a0is a record-and-replay framework which tests a given workload on the target file system by\u00a0<i>simulating<\/i>\u00a0power-loss crashes while the workload is being executed, and checking if the file system recovers to a correct state after each crash. A<span class=\"gs_fscp\">ce<\/span>\u00a0automatically generates all the workloads to be run on the target file system. We build C<span class=\"gs_fscp\">rash<\/span>M<span class=\"gs_fscp\">onkey<\/span>\u00a0and A<span class=\"gs_fscp\">ce<\/span>\u00a0based on a new approach to test file-system crash consistency:\u00a0<i>bounded black-box crash testing<\/i>\u00a0(<i>B<\/i><sup>3<\/sup>).\u00a0<i>B<\/i><sup>3<\/sup>\u00a0tests the file system in a black-box manner using workloads of file-system operations. Since the space of possible workloads is infinite,\u00a0<i>B<\/i><sup>3<\/sup> bounds this space based on parameters such as the number of file-system operations or which operations to include, and exhaustively generates workloads within this bounded space. <i>B<\/i><sup>3<\/sup> builds upon insights derived from our study of crash-consistency bugs reported in Linux file systems in the last 5 years. We observed that most reported bugs can be reproduced using small workloads of three or fewer file-system operations on a newly created file system, and that all reported bugs result from crashes after fsync()-related system calls. CrashMonkey and Ace are able to find 24 out of the 26 crash-consistency bugs reported in the last 5 years. Our tools also revealed 10 new crash-consistency bugs in widely used, mature Linux file systems, 7 of which existed in the kernel since 2014. Additionally, our tools found a crash-consistency bug in a verified file system, FSCQ. The new bugs result in severe consequences like broken rename atomicity, loss of persisted files and directories, and data loss.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We present CrashMonkey\u00a0and Ace, a set of tools to systematically find crash-consistency bugs in Linux file systems. CrashMonkey\u00a0is a record-and-replay framework which tests a given workload on the target file system by\u00a0simulating\u00a0power-loss crashes while the workload is being executed, and checking if the file system recovers to a correct state after each crash. Ace\u00a0automatically generates [&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":"","msr-author-ordering":null,"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"ACM Transactions on Storage (TOS)","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":"","msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2019-4-20","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":0,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":false,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13547],"msr-publication-type":[193715],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-788258","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2019-4-20","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"ACM Transactions on Storage (TOS)","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":0,"msr_main_download":"","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2021\/10\/tos-crashmonkey.pdf","id":"788261","title":"tos-crashmonkey","label_id":"243109","label":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[{"id":788261,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2021\/10\/tos-crashmonkey.pdf"}],"msr-author-ordering":[{"type":"user_nicename","value":"Jayashree Mohan","user_id":40927,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Jayashree Mohan"},{"type":"text","value":"Ashlie Martinez","user_id":0,"rest_url":false},{"type":"text","value":"Soujanya Ponnapalli","user_id":0,"rest_url":false},{"type":"text","value":"Pandian Raju","user_id":0,"rest_url":false},{"type":"text","value":"Vijay Chidambaram","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"article","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/788258","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":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/788258\/revisions"}],"predecessor-version":[{"id":788264,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/788258\/revisions\/788264"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=788258"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=788258"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=788258"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=788258"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=788258"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=788258"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=788258"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=788258"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=788258"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=788258"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=788258"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=788258"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=788258"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}