On Thu, Feb 21, 2019 at 03:22:56PM -0300, Fabio Abreu wrote: :Hi Everybody, : :It's recommended join different hardwares in the same rack ? This is based on my somewhat subjective experience not rigorous testing or deep understanding of the code base so your results may vary but ... Physical location probably doesn't matter too much here, but pool performance is generally limitted by the lowest performing set of disks in the pool (it's a bit more complex but that's my experience). I'm vioulating that recommendation now as I replace older HDD with SSD. We've written crush rules to require one replica on ssd and the rest on HDD for teh mixed pools and then set primary affinity such that the SSD are always thw primary replica, this enhnaces read performace and the HDDs are already using SSD WAL so writes are also all SSD. But that's a very specific use case and long term were are going all SSD just not all at once. If you have one set of hardware that's causing performance problems it's probably best to create a pool that is all the low performance disk and try to use that for less demanding applications if you can. -Jon :For example I have a sata rack with Apollo 4200 storage and I will get :another hardware type to expand this rack, Hp 380 Gen10. : :I was made a lot tests to understand the performance and these new disks :have 100% of utilization in my environment and the cluster recovery is :worst than another hardware. : :Can Someone recommend a best practice or configuration in this scenario? I :make this issue because if these disks not performing as hope, I will :configure another pools to my openstack and maybe that's not make sense to :me because I will split nova process in the computes node if I have two :pools . : : :Regards, :Fabio Abreu :_______________________________________________ :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