Hi,
maybe you can help us with following probs:
if you need more info about our cluster or any debugging log I will be happy to help
Environment:
-------------------
Small test cluster with 7 node, 1 osd per node
Upgrade from dumpling to emporer 0.72.1
2 Problems after upgrade:
-----------------------------------
- ceph -s shows <health HEALTH_WARN pool images has too few pgs>
- changing pg_num etc from 1024 to 1600 has no effect
- after setting <mon_pg_warn_min_objects> to a larger value the cluster moves to HEALTH_OK
- 20 pools
create new pool
wait until pool is ready
delete pool again
ceph -s still shows 21 pools
stop first mon, wait, start mon again
ceph -s shows 20 pools, 21 pools, 20 pools, ... changing every n (~ 2) secs
stop second mon, wait, start mon again
ceph -s shows 20 pools, 21 pools, 20 pools, ... changing every m > n secs
stop third mon, wait, start mon again
ceph -s shows 20 pools ... stable
Some infos about the cluster:
----------------------------------------
# uname –a
Linux host1 3.2.0-53-generic #81-Ubuntu SMP Thu Aug 22 21:01:03 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
# ceph --version
ceph version 0.72.1 (4d923861868f6a15dcb33fef7f50f674997322de)
# ceph -s
cluster 3f4a289a-ad40-40e7-8204-ab38affb18f8
health HEALTH_WARN pool images has too few pgs
monmap e1: 3 mons at {host1=10.255.NNN.NN:6789/0,host2=10.255.NNN.NN:6789/0,host3=10.255.NNN.NN:6789/0},\
election epoch 306, quorum 0,1,2 host1,host2,host3
mdsmap e21: 1/1/1 up {0=host4=up:active}
osdmap e419: 7 osds: 7 up, 7 in
pgmap v815042: 17152 pgs, 21 pools, 97795 MB data, 50433 objects
146 GB used, 693 GB / 839 GB avail
17152 active+clean
client io 25329 kB/s wr, 39 op/s
# ceph osd dump
epoch 419
fsid 3f4a289a-ad40-40e7-8204-ab38affb18f8
created 2013-09-27 14:37:13.601605
modified 2013-11-20 16:55:35.791740
flags
pool 0 'data' rep size 2 min_size 1 crush_ruleset 0 object_hash rjenkins pg_num 64 pgp_num 64 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 64 pgp_num 64 last_change 1 owner 0
pool 2 'rbd' rep size 2 min_size 1 crush_ruleset 2 object_hash rjenkins pg_num 64 pgp_num 64 last_change 1 owner 0
pool 3 'images' rep size 2 min_size 1 crush_ruleset 0 object_hash rjenkins pg_num 1600 pgp_num 1600 last_change 367 owner 0
removed_snaps [1~1,3~2]
...
...
# rados lspools | wc -l
20
Greetings
Michael
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
