thanks a lot to your reply. i know that [5,3], osd.5 is the primary osd, since my replicate size is 2. and in my testing cluster. test.txt only have this only one file. i just mount -t cephfs 192.168.250.15:6789:/ , so means ,use pool data by default ? ##The acting OSDs however are the OSD numbers. So for this PG 5 is primary and 3 secondary.## i am not understand it . up[5,3] means osd.5 is primary,osd.3 is replicate. acting[5,3] what is the meaning? my purpose just want to know how is mapping between object and the file. i read some information from: http://www.ibm.com/developerworks/library/l-ceph/ , that is why some confuse about it. ##Rather than rely on allocation lists (metadata to map blocks on a disk to a given file), Ceph uses an interesting alternative. A file from the Linux perspective is assigned an inode number (INO) from the metadata server, which is a unique identifier for the file. The file is then carved into some number of objects (based on the size of the file). Using the INO and the object number (ONO), each object is assigned an object ID (OID). Using a simple hash over the OID, each object is assigned to a placement group. The placement group (identified as a PGID) is a conceptual container for objects. Finally, the mapping of the placement group to object storage devices is a pseudo-random mapping using an algorithm called Controlled Replication Under Scalable Hashing(CRUSH). In this way, mapping of placement groups (and replicas) to storage devices does not rely on any metadata but instead on a pseudo-random mapping function. This behavior is ideal, because it minimizes the overhead of storage and simplifies the distribution and lookup of data## ========== your last reply ========== On 08/15/2013 05:18 PM wido wrote: > mount cephfs to /mnt/mycehfs on debian 7, kernel3.10 > > eg: have one file > root@test-debian:/mnt/mycephfs# ls -i test.txt > 1099511627776 test.txt > root@test-debian:/mnt/mycephfs# ceph osd map volumes test.txt So you used the pool volumes here when mounting instead of the pool "data" ? > osdmap e351 pool 'volumes' (3) object 'test.txt' -> pg 3.8b0b6108 (3.8) > -> up [5,3] acting [5,3] I don't think it's that easy since you also have to calculate in the path name of the file. I haven't done this myself, but your method would imply there can only be one "test.txt" in that whole pool. The acting OSDs however are the OSD numbers. So for this PG 5 is primary and 3 secondary. > root@test-debian:/mnt/mycephfs# ceph osd pool get volumes size > size: 2 > root@test-debian:/mnt/mycephfs# ceph osd stat > e351: 7 osds: 7 up, 7 in > root@test-debian:/mnt/mycephfs# ceph osd dump > > epoch 351 > fsid db32486a-7ad3-4afe-8b67-49ee2a6dcecf > created 2013-08-08 13:45:52.579015 > modified 2013-08-15 02:39:45.473969 > flags > > pool 0 'data' rep size 2 min_size 1 crush_ruleset 0 object_hash rjenkins > pg_num 192 pgp_num 192 last_change 1 owner 0 crash_replay_interval 45 > pool 1 'metadata' rep size 2 min_size 1 crush_ruleset 1 object_hash > rjenkins pg_num 192 pgp_num 192 last_change 1 owner 0 > pool 2 'rbd' rep size 2 min_size 1 crush_ruleset 2 object_hash rjenkins > pg_num 192 pgp_num 192 last_change 1 owner 0 > pool 3 'volumes' rep size 2 min_size 1 crush_ruleset 0 object_hash > rjenkins pg_num 256 pgp_num 256 last_change 220 owner 18446744073709551615 > > > how to know where is store in ceph,which object, which pg? i want > understand how ceph work. > In your example the data is on osd.3 and osd.5 Wido > pls give your guide. > > best wish. _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |