Hi Alfredo, you may want to check the SMART data for the disk. I also had such a case recently (see http://lists.ceph.com/pipermail/ceph-users-ceph.com/2019-May/035117.html for the thread), and the disk had one unreadable sector which was pending reallocation. Triggering "ceph pg repair" for the problematic placement group made the OSD rewrite the problematic sector and allowed the disk to reallocate this unreadable sector. Cheers, Oliver Am 06.06.19 um 18:45 schrieb Tarek Zegar:
Look here _http://docs.ceph.com/docs/master/rados/troubleshooting/troubleshooting-pg/#pgs-inconsistent_ Read error typically is a disk issue. The doc is not clear on how to resolve that Inactive hide details for Alfredo Rezinovsky ---06/06/2019 10:58:50 AM---https://urldefense.proofpoint.com/v2/url?u=https-3A__cAlfredo Rezinovsky ---06/06/2019 10:58:50 AM---https://urldefense.proofpoint.com/v2/url?u=https-3A__ceph.com_geen-2Dcategorie_ceph-2Dmanually-2Drep From: Alfredo Rezinovsky <alfrenovsky@xxxxxxxxx> To: Ceph Users <ceph-users@xxxxxxxxxxxxxx> Date: 06/06/2019 10:58 AM Subject: [EXTERNAL] Fix scrub error in bluestore. Sent by: "ceph-users" <ceph-users-bounces@xxxxxxxxxxxxxx> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ _https://ceph.com/geen-categorie/ceph-manually-repair-object/_ is a little outdated. After stopping the OSD, flushing the journal I don't have any clue on how to move the object (easy in filestore). I have thins in my osd log. 2019-06-05 10:46:41.418 7f47d0502700 -1 log_channel(cluster) log [ERR] : 10.c5 shard 2 soid 10:a39e2c78:::1000083f81f.00000001:head : candidate had a read error How can I fix it? -- Alfrenovsky_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
