Alright, I was finally able to get this resolved without adding another node. As pointed out, even though I had a config variable that defined the default replicated size at 2, ceph for some reason created the default pools (data, and metadata) with a value of 3. After digging trough documentation I found: ceph osd dump | grep 'replicated size' Which shows the replicated size for each pool. My newly created pools ssd and sata were correctly configured, but the default pools in ceph were not. I was then able to set: ceph osd pool set metadata size 2 and ceph osd pool set data size 2 Finally, my cluster is healthy! Not exactly straight forward installation and troubleshooting, but it works. Thanks for the help and tips along the way. The advice definitely led me in the right direction.
 |