CRUSH Rule Map

rule replicated_ruleset {
        ruleset 0
        type replicated
        min_size 1
        max_size 10
        step take default
        step choose firstn 0 type host
        step chooseleaf firstn -1 type osd
        step emit
}


2 hosts
5 osds per host
Total of 10 osds.
Pool size 3

If I'm understanding this correctly, when trying to find a location for a pg, it should first select 3 hosts (of which it should reduce to 2 as there are only 2 hosts), then for each host, try and select 2 osds (of which, on the second host, it only selects 1 due to 3 replicas).  This should have a net result of always placing two replicas on one host and 1 replica on the other host.  The primary should also always have a second copy on the same host.