I finally have my first test cluster up and running. No data on it, yet. The config is: three mons, and three OSDS servers. Each OSDS server has eight 4TB SAS drives and two SSD journal drives. The cluster is healthy, so I started playing with PG and PGP values. By the provided calculation is determined that 600 PG/PGP was the recommended value, so I set both to 600 (replication of 3). I decided that 600 might be too high, so I attempted to lower it to 512 (next power-of-2 number from 600). I, then, realized that I couldn’t lower it, only raise it from the present value. So, I deleted the pools that I changed (rbd and data). Then, I recreated the pools, but noticed that it created the two pools with pool numbers that were different from what it started with. My two questions are: is there any way to recreate the pools with the original pool number value that they had? Does it really matter that data pool is now 3, and rbd is now 4? Any other recommendations that the list can provide regarding PG’s and PGP’s?
|
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
