Hi everyone, I am new to Ceph and trying to test out my understanding on the CRUSH map. Attached is a hypothetical cluster diagram with 3 racks. On each rack, the first host runs 3 SSD-based OSDs and the second 3 HDD-based. My goal is to create two rules that separate SSD and HDD performance domains (by using device class) and both rules should use a*rack* level failure domain.
Please, provide on human language what actually hosts configuration you have and what drive pattern usage you want at finish?
Obviously you want just separate your ssd/hdd load per pool basis, but I need more information to clarify.
k _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
