Strahil Nikolov
2022-Dec-13 13:45 UTC
[Gluster-users] Fwd: really large number of skipped files after a scrub
Based on?https://bugzilla.redhat.com/show_bug.cgi?id=1299737#c12 , the previos name was 'number of unsigned files'. Signing seem to be a very complex process (see http://goo.gl/Mjy4mD ) and as far as I understand - those 'skipped' files were too new to be signed . If you do have RAID5/6 , I think that bitrod is unnecessary. Best Regards,Strahil Nikolov? On Tue, Dec 13, 2022 at 12:33, cYuSeDfZfb cYuSeDfZfb<cyusedfzfb at gmail.com> wrote: Hi, I am running a PoC with cluster, and, as one does, I am trying to break and heal it. One of the things I am testing is scrubbing / healing. My cluster is created on ubuntu 20.04 with stock glusterfs 7.2, and my test volume info: Volume Name: gv0 Type: Replicate Volume ID: 7c09100b-8095-4062-971f-2cea9fa8c2bc Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: gluster1:/data/brick1/gv0 Brick2:?gluster2:/data/brick1/gv0 Brick3:?gluster3:/data/brick1/gv0 Options Reconfigured: features.scrub-freq: daily auth.allow: x.y.z.q transport.address-family: inet storage.fips-mode-rchecksum: on nfs.disable: on performance.client-io-threads: off features.bitrot: on features.scrub: Active features.scrub-throttle: aggressive storage.build-pgfid: on I have two issues: 1) scrubs are configured to run daily (see above) but they don't automatically happen. Do I need to configure something to actually get daily automatic scrubs? 2) A "scrub status" reports *many* skipped files, and only very few files that have actually been scrubbed. Why are so many files skipped? See: gluster volume bitrot gv0 scrub status Volume name : gv0 State of scrub: Active (Idle) Scrub impact: aggressive Scrub frequency: daily Bitrot error log location: /var/log/glusterfs/bitd.log Scrubber error log location: /var/log/glusterfs/scrub.log ======================================================== Node: localhost Number of Scrubbed files: 8112 Number of Skipped files: 51209 Last completed scrub time: 2022-12-10 04:36:55 Duration of last scrub (D:M:H:M:S): 0:16:58:53 Error count: 0 ======================================================== Node:?gluster3 Number of Scrubbed files: 42 Number of Skipped files: 59282 Last completed scrub time: 2022-12-10 02:24:42 Duration of last scrub (D:M:H:M:S): 0:16:58:15 Error count: 0 ======================================================== Node:?gluster2 Number of Scrubbed files: 42 Number of Skipped files: 59282 Last completed scrub time: 2022-12-10 02:24:29 Duration of last scrub (D:M:H:M:S): 0:16:58:2 Error count: 0 ========================================================Thanks!MJ ________ Community Meeting Calendar: Schedule - Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC Bridge: https://meet.google.com/cpu-eiue-hvk Gluster-users mailing list Gluster-users at gluster.org https://lists.gluster.org/mailman/listinfo/gluster-users -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20221213/8bca5469/attachment.html>
Strahil Nikolov
2022-Dec-13 13:48 UTC
[Gluster-users] Fwd: really large number of skipped files after a scrub
By the way, what is the output of 'ps aux | grep bitd' ? Best Regards,Strahil Nikolov? On Tue, Dec 13, 2022 at 15:45, Strahil Nikolov<hunter86_bg at yahoo.com> wrote: Based on?https://bugzilla.redhat.com/show_bug.cgi?id=1299737#c12 , the previos name was 'number of unsigned files'. Signing seem to be a very complex process (see http://goo.gl/Mjy4mD ) and as far as I understand - those 'skipped' files were too new to be signed . If you do have RAID5/6 , I think that bitrod is unnecessary. Best Regards,Strahil Nikolov? On Tue, Dec 13, 2022 at 12:33, cYuSeDfZfb cYuSeDfZfb<cyusedfzfb at gmail.com> wrote: Hi, I am running a PoC with cluster, and, as one does, I am trying to break and heal it. One of the things I am testing is scrubbing / healing. My cluster is created on ubuntu 20.04 with stock glusterfs 7.2, and my test volume info: Volume Name: gv0 Type: Replicate Volume ID: 7c09100b-8095-4062-971f-2cea9fa8c2bc Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: gluster1:/data/brick1/gv0 Brick2:?gluster2:/data/brick1/gv0 Brick3:?gluster3:/data/brick1/gv0 Options Reconfigured: features.scrub-freq: daily auth.allow: x.y.z.q transport.address-family: inet storage.fips-mode-rchecksum: on nfs.disable: on performance.client-io-threads: off features.bitrot: on features.scrub: Active features.scrub-throttle: aggressive storage.build-pgfid: on I have two issues: 1) scrubs are configured to run daily (see above) but they don't automatically happen. Do I need to configure something to actually get daily automatic scrubs? 2) A "scrub status" reports *many* skipped files, and only very few files that have actually been scrubbed. Why are so many files skipped? See: gluster volume bitrot gv0 scrub status Volume name : gv0 State of scrub: Active (Idle) Scrub impact: aggressive Scrub frequency: daily Bitrot error log location: /var/log/glusterfs/bitd.log Scrubber error log location: /var/log/glusterfs/scrub.log ======================================================== Node: localhost Number of Scrubbed files: 8112 Number of Skipped files: 51209 Last completed scrub time: 2022-12-10 04:36:55 Duration of last scrub (D:M:H:M:S): 0:16:58:53 Error count: 0 ======================================================== Node:?gluster3 Number of Scrubbed files: 42 Number of Skipped files: 59282 Last completed scrub time: 2022-12-10 02:24:42 Duration of last scrub (D:M:H:M:S): 0:16:58:15 Error count: 0 ======================================================== Node:?gluster2 Number of Scrubbed files: 42 Number of Skipped files: 59282 Last completed scrub time: 2022-12-10 02:24:29 Duration of last scrub (D:M:H:M:S): 0:16:58:2 Error count: 0 ========================================================Thanks!MJ ________ Community Meeting Calendar: Schedule - Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC Bridge: https://meet.google.com/cpu-eiue-hvk Gluster-users mailing list Gluster-users at gluster.org https://lists.gluster.org/mailman/listinfo/gluster-users -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20221213/199fc3f6/attachment.html>