Hello,
i'm facing a problem that mds seems not to start.
I started mds in debug mode "ceph-mds -f -i storage08 --debug_mds 10" which outputs in the log:
---------------------- cut ---------------------------------
2015-09-21 14:12:14.313534 7ff47983d780 0 ceph version 0.94.3 (95cefea9fd9ab740263bf8bb4796fd864d9afe2b), process ceph-mds, pid 24787
starting mds.storage08 at :/0
2015-09-21 14:12:14.316062 7ff47983d780 10 mds.-1.0 168 MDSCacheObject
2015-09-21 14:12:14.316074 7ff47983d780 10 mds.-1.0 2408 CInode
2015-09-21 14:12:14.316075 7ff47983d780 10 mds.-1.0 16 elist<>::item *7=112
2015-09-21 14:12:14.316077 7ff47983d780 10 mds.-1.0 480 inode_t
2015-09-21 14:12:14.316079 7ff47983d780 10 mds.-1.0 48 nest_info_t
2015-09-21 14:12:14.316081 7ff47983d780 10 mds.-1.0 32 frag_info_t
2015-09-21 14:12:14.316082 7ff47983d780 10 mds.-1.0 40 SimpleLock *5=200
2015-09-21 14:12:14.316083 7ff47983d780 10 mds.-1.0 48 ScatterLock *3=144
2015-09-21 14:12:14.316085 7ff47983d780 10 mds.-1.0 480 CDentry
2015-09-21 14:12:14.316086 7ff47983d780 10 mds.-1.0 16 elist<>::item
2015-09-21 14:12:14.316096 7ff47983d780 10 mds.-1.0 40 SimpleLock
2015-09-21 14:12:14.316097 7ff47983d780 10 mds.-1.0 952 CDir
2015-09-21 14:12:14.316098 7ff47983d780 10 mds.-1.0 16 elist<>::item *2=32
2015-09-21 14:12:14.316099 7ff47983d780 10 mds.-1.0 176 fnode_t
2015-09-21 14:12:14.316100 7ff47983d780 10 mds.-1.0 48 nest_info_t *2
2015-09-21 14:12:14.316101 7ff47983d780 10 mds.-1.0 32 frag_info_t *2
2015-09-21 14:12:14.316103 7ff47983d780 10 mds.-1.0 264 Capability
2015-09-21 14:12:14.316104 7ff47983d780 10 mds.-1.0 32 xlist<>::item *2=64
2015-09-21 14:12:14.316665 7ff47983d780 -1 mds.-1.0 log_to_monitors {default=true}
2015-09-21 14:12:14.320840 7ff4740c8700 7 mds.-1.server handle_osd_map: full = 0 epoch = 20
2015-09-21 14:12:14.320984 7ff47983d780 10 mds.beacon.storage08 _send up:boot seq 1
2015-09-21 14:12:14.321060 7ff47983d780 10 mds.-1.0 create_logger
2015-09-21 14:12:14.321234 7ff4740c8700 5 mds.-1.0 handle_mds_map epoch 1 from mon.1
2015-09-21 14:12:14.321256 7ff4740c8700 10 mds.-1.0 my compat compat={},rocompat={},incompat={1=base v0.20,2=client writeable ranges,3=default file layouts on dirs,4=dir inode in separate object,5=mds uses versioned encoding,6=dirfrag is stored in omap,7=mds uses inline data,8=no anchor table}
2015-09-21 14:12:14.321264 7ff4740c8700 10 mds.-1.0 mdsmap compat compat={},rocompat={},incompat={}
2015-09-21 14:12:14.321267 7ff4740c8700 10 mds.-1.-1 map says i am 192.168.0.178:6802/24787 mds.-1.-1 state down:dne
2015-09-21 14:12:14.321272 7ff4740c8700 10 mds.-1.-1 not in map yet
2015-09-21 14:12:14.321305 7ff4740c8700 7 mds.-1.server handle_osd_map: full = 0 epoch = 20
2015-09-21 14:12:14.321443 7ff4740c8700 5 mds.-1.-1 handle_mds_map epoch 1 from mon.1
2015-09-21 14:12:14.321447 7ff4740c8700 5 mds.-1.-1 old map epoch 1 <= 1, discarding
2015-09-21 14:12:18.321061 7ff4707c0700 10 mds.beacon.storage08 _send up:boot seq 2
2015-09-21 14:12:19.321093 7ff470fc1700 10 MDSInternalContextBase::complete: N3MDS10C_MDS_TickE
2015-09-21 14:12:22.321119 7ff4707c0700 10 mds.beacon.storage08 _send up:boot seq 3
2015-09-21 14:12:24.321169 7ff470fc1700 10 MDSInternalContextBase::complete: N3MDS10C_MDS_TickE
...
---------------------- cut ---------------------------------
"cheph -s" shows:
cluster 982924a3-32e7-401f-9975-018bb697d717
health HEALTH_OK
monmap e1: 3 mons at {0=192.168.0.176:6789/0,1=192.168.0.177:6789/0,2=192.168.0.178:6789/0}
election epoch 6, quorum 0,1,2 0,1,2
osdmap e20: 3 osds: 3 up, 3 in
pgmap v39: 64 pgs, 1 pools, 0 bytes data, 0 objects
15541 MB used, 388 GB / 403 GB avail
64 active+clean
As you see MONs and OSDs seem to be happy.
I'm missing the "mdsmap" entry here. Try to verify with the command "ceph mds stat" gives:
e1: 0/0/0 up
The section of ceph.conf regarding mds reads:
[mds]
mds data = /var/lib/ceph/mds/ceph-$id
keyring = /var/lib/ceph/mds/ceph-$id/keyring
[mds.storage08]
host = storage08
mds addr = 192.168.0.178
Configuration:
3 Hosts
Gentoo Linux (kernel 4.0.5)
Ceph 0.94.3
All have MON and OSD
One has MDS additionally.
Any idea for on how the MDS to get running ?
Kind regards
Petric
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
