NVMe (hope those are enterprise not client) drives aren't likely to suffer the same bottlenecks as HDDs or even SATA SSDs. And a 2:1 size ratio isn't the largest I've seen. So I would just use all 108 OSDs as a single device class and spread the pools across all of them. That way you won't run out of space in one before the other. When running OSDs of multiple sizes I do recommend setting mon_max_pg_per_osd to 1000 so that the larger ones won't run afoul of the default value, especially when there's a failure. > > Hi, > > I need some guidance from you folks... > > I am going to deploy a ceph cluster in HCI mode for an openstack platform. > My hardware will be : > - 03 control nodes : > - 27 osd nodes : each node has 03x3.8To nvme + 01x1.9To nvme disks (those > disks will all be used as OSDs) > > In my Openstack I will be creating all sorts of pools : RBD, Cephfs and RGW. > > I am planning to create two crush rules using the disk size as a parameter. > Then divide my pools between the two rules. > - RBD to use the 3.8To disks since I need more space here. > - Cephfs and RGW to use 1.9To disks. > > Is this a good configuration? > > Regards > _______________________________________________ > 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
 |