Hi All, I am facing a weird situation with Ceph Block device (rbd) -
The blkio.weight feature of cgroup seems to be not working on Ceph block device (say on /dev/rbd1 and /dev/rbd1), the iops number reported by the workload (fio or dd) should be in proportion to the applied weights (weights applied to VM or to a workload tasks). Though blkio.weight feature of cgroup is working for normal device (ex- sda, sdb). I am setting IO schedulers to CFQ for both rbd and normal disks. Any clue what i am missing. Regards, Vikrant On Mon, Mar 10, 2014 at 12:08 AM, Vikrant Verma <vikrantverm...@gmail.com>wrote: > Hi All, > > blkio weight sharing is not working on Ceph Block device (RBD),IOPS > number are not getting divided as per the weight proportion but are getting > divided equally. > i am using CFQ IO scheduler for RBD. > > I did it getting it worked once but now i am not able to get it worked as > i have upgraded ceph and ceph-deploy version of the cluster. > > Is there any dependency of blkio weight on ceph release. > > Please let me know if you have any idea about it. Early response will be > highly appreciated. > > Regards, > Vikrant >
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
