I have fixed incomplete PGs in my environment. And i believe the situation was cuased by OSD flaps during backfilling. So i am asking for general guidelines on 1) how to avoid incomplete PGs as much as possible, since it risks data loss; 2) is there a tool or script to reliably fix it? best regards, samuel huxiaoyu@xxxxxxxxxxxx From: Eugen Block Date: 2020-08-15 12:51 To: huxiaoyu CC: ceph-users Subject: Re: how to handle incomplete PGs Please provide more details about your environment, otherwise it's just guessing what could have happened. Zitat von huxiaoyu@xxxxxxxxxxxx: > I am using replicated pool, and min_size=1. I do not have any disk > failure, so i do not expect incomplete PGs, but it appeared after > OSD flaped. > > > > huxiaoyu@xxxxxxxxxxxx > > From: Eugen Block > Date: 2020-08-15 09:39 > To: huxiaoyu > CC: ceph-users > Subject: Re: how to handle incomplete PGs > Hi, > > did you wait for the backfill to complete before removing the old > drives? What is your environment? Are the affected PGs from an EC > pool? Does [1] apply to you? > > [1] http://lists.ceph.com/pipermail/ceph-users-ceph.com/2019-July/035743.html > > > Zitat von huxiaoyu@xxxxxxxxxxxx: > >> Dear Ceph folks, >> >> Recently i encountered incomplete PGs when replacing an OSD node >> with new handware. I noticed multiple OSD ups and downs, and >> eventually a few PGs got stucked at PG incomplete status. >> >> Questions 1: is there a reliable way to avoid the occurence of >> incomplete PGs? >> 2: is there a good tool or scriptes to handle >> incomplete PGs without lossing data >> >> best regards, >> >> samuel >> >> >> >> huxiaoyu@xxxxxxxxxxxx >> _______________________________________________ >> ceph-users mailing list -- ceph-users@xxxxxxx >> To unsubscribe send an email to ceph-users-leave@xxxxxxx _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |