On a recent Luminous cluster, with nvme*n1 devices, the class is automatically set as "nvme" on "Intel SSD DC P3520 Series" : ~# ceph osd tree ID CLASS WEIGHT TYPE NAME STATUS REWEIGHT PRI-AFF -1 2.15996 root default -9 0.71999 room room135 -3 0.71999 host ceph135a 0 nvme 0.35999 osd.0 up 1.00000 1.00000 1 nvme 0.35999 osd.1 up 1.00000 1.00000 -11 0.71999 room room209 -5 0.71999 host ceph209a 2 nvme 0.35999 osd.2 up 1.00000 1.00000 3 nvme 0.35999 osd.3 up 1.00000 1.00000 -12 0.71999 room room220 -7 0.71999 host ceph220a 4 nvme 0.35999 osd.4 up 1.00000 1.00000 5 nvme 0.35999 osd.5 up 1.00000 1.00000 Le dimanche 12 août 2018 à 23:37 +0200, ceph@xxxxxxxxxx a écrit : > > Am 1. August 2018 10:33:26 MESZ schrieb Jake Grimmett < > jog@xxxxxxxxxxxxxxxxx>: > > Dear All, > > Hello Jake, > > > > > Not sure if this is a bug, but when I add Intel Optane 900P drives, > > their device class is automatically set to SSD rather than NVME. > > > > AFAIK ceph actually difference only between hdd and ssd. Nvme would > be handled as same like ssd. > > Hth > - Mehmet > > > This happens under Mimic 13.2.0 and 13.2.1 > > > > [root@ceph2 ~]# ceph-volume lvm prepare --bluestore --data > > /dev/nvme0n1 > > > > (SNIP see http://p.ip.fi/eopR for output) > > > > Check... > > [root@ceph2 ~]# ceph osd tree | grep "osd.1 " > > 1 ssd 0.25470 osd.1 up 1.00000 1.00000 > > > > Fix is easy > > [root@ceph2 ~]# ceph osd crush rm-device-class osd.1 > > done removing class of osd(s): 1 > > > > [root@ceph2 ~]# ceph osd crush set-device-class nvme osd.1 > > set osd(s) 1 to class 'nvme' > > > > Check... > > [root@ceph2 ~]# ceph osd tree | grep "osd.1 " > > 1 nvme 0.25470 osd.1 up 1.00000 1.00000 > > > > > > Thanks, > > > > Jake > > _______________________________________________ > > 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 > _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |