Now I'm building another ceph cluster with 0.60, but I'm still running v0.55.1 in production. I used to set 'ceph osd reweight' value same as 'ceph osd crush reweight', for example weight 3 for 3TB harddisk, but it's impossible in v0.60, it ranges 0..1 now. So, what changed? In my opinion, in v0.60, 'ceph osd crush reweight' makes ceph distribute data across OSDs by their weight, and the 'ceph osd reweight' controls migration speed of data in ceph cluster. Right ? # id weight type name up/down reweight -1 92 root default -3 92 rack unknownrack -2 26 host c15 0 3 osd.0 up 1 1 3 osd.1 up 1 10 2 osd.10 up 1 2 3 osd.2 up 1 3 3 osd.3 up 1 4 2 osd.4 up 1 5 2 osd.5 up 1 6 2 osd.6 up 1 7 2 osd.7 up 1 8 2 osd.8 up 1 9 2 osd.9 up 1 -4 33 host c16 11 3 osd.11 up 1 12 3 osd.12 up 1 13 3 osd.13 up 1 14 3 osd.14 up 1 15 3 osd.15 up 1 16 3 osd.16 up 1 17 3 osd.17 up 1 18 3 osd.18 up 1 19 3 osd.19 up 1 20 3 osd.20 up 1 21 3 osd.21 up 1 -5 33 host c18 22 3 osd.22 up 1 23 3 osd.23 up 1 24 3 osd.24 up 1 25 3 osd.25 up 1 26 3 osd.26 up 1 27 3 osd.27 up 1 28 3 osd.28 up 1 29 3 osd.29 up 1 30 3 osd.30 up 1 31 3 osd.31 up 1 32 3 osd.32 up 1 -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
