The goal should be to increase the weights in unison, which should prevent any actual data movement (modulo some rounding error, perhaps). At the moment that can't be done via the CLI, but you can: ceph osd getcrusshmap -o /tmp/cm crushtool -i /tmp/cm --reweight-item osd.0 3.5 --reweight-item osd.1 3.5 ... -o /tmp/cm2 crushtool -d /tmp/cm2 # make sure it looks right ceph osd setcrushmap -i /tmp/cm2 On Tue, 4 Mar 2014, Udo Lembke wrote: > Hi all, > I have startet the ceph-cluster with an weight of 1 for all osd-disks (4TB). > Later I switched to ceph-deploy and ceph-deploy use normaly an weight of > 3.64 for this disks, which makes much more sense! > > Now I wan't to change the weight of all 52 osds (on 4 nodes) to 3.64 and > the question is, how to proceed on an productional cluster? > > Increase all 52 weights in 0.1-steps, wait for an calm system and do it > again till 3.64 is reached? > Or modify the crushmap (export, decompile, change, compile, load) > directly from 1 to 3.64? Is in this case the data accessible? > Are OSDs with 76% data an problem in this scenario? I have learned, that > ceph really don't like full disks ;-) > > Any hints? > > > Udo > _______________________________________________ > 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
 |