On Fri, Jan 17, 2014 at 8:41 AM, Wido den Hollander <wido@xxxxxxxx> wrote: > On 01/17/2014 01:29 PM, Joao Eduardo Luis wrote: >> >> On 01/17/2014 12:02 PM, Wido den Hollander wrote: >>> >>> On 01/17/2014 12:46 PM, Tim Zhang wrote: >>>> >>>> Hi guys, >>>> I use ceph-deploy to deploy my ceph cluster. >>>> This is my config file: >>>> >>>> ----------------------------------------------------------------------------- >>>> >>>> >>>> [global] >>>> osd pool default size = 3 >>>> auth_service_required = none >>>> filestore_xattr_use_omap = true >>>> journal zero on create = true >>>> auth_client_required = none >>>> auth_cluster_required = none >>>> mon_host = 192.168.1.172,192.168.1.130,192.168.1.115 >>>> osd_journal_size = 1024 >>>> public_network = 192.168.1.0/24 <http://192.168.1.0/24> >>>> mon_initial_members = node30, node31, node32 >>>> cluster_network = 192.168.1.0/24 <http://192.168.1.0/24> >>>> fsid = da79afb2-d85e-406a-b05b-80eaaac2e179 >>>> >>>> ----------------------------------------------------------------------------- >>>> >>>> >>>> >>>> after deploying , the cluster is unhealthy and I find mon addr's is not >>>> as the config file setting, and the dedicated mon addrs(mon_host ) are >>>> in the extra_probe_peers fields according to the command: >>>> ceph --cluster=ceph --admin-daemon /var/run/ceph/ceph-mon.node30.asok >>>> mon_status; >>>> and seems the cluster using the other ip addr on the hosts. Thats >>>> strange, can anyone give some suggestions? >>>> >>> >>> What version of Ceph is this? I've seen something similar with Dumpling >>> where the mon would bind to 0.0.0.0:6800. I haven't been able to find >>> the root cause yet. >>> >>> Wido >> >> >> http://tracker.ceph.com/issues/5804 >> >> Any info on how to reproduce this thing is welcome. >> > > Done. I added a comment to the issue regarding how I saw it happen. I can > reproduce it any moment using those VirtualBox images. > > Wido > > >> -Joao >> >>> >>>> the output is as following: This input is not the full output that would come out from ceph-deploy. One of the useful things that ceph-deploy does is that it tries to be very granular as to what is going on so that debugging is easier. Would you mind pasting the complete log output and maybe add it to the ticket as well? >>>> [node30][INFO ] Running command: ceph --cluster=ceph --admin-daemon >>>> /var/run/ceph/ceph-mon.node30.asok mon_status >>>> [node30][DEBUG ] >>>> >>>> ******************************************************************************** >>>> >>>> >>>> [node30][DEBUG ] status for monitor: mon.node30 >>>> [node30][DEBUG ] { >>>> [node30][DEBUG ] "election_epoch": 0, >>>> [node30][DEBUG ] "extra_probe_peers": [ >>>> [node30][DEBUG ] "192.168.1.115:6789/0 >>>> <http://192.168.1.115:6789/0>", >>>> [node30][DEBUG ] "192.168.1.130:6789/0 >>>> <http://192.168.1.130:6789/0>", >>>> [node30][DEBUG ] "192.168.1.172:6789/0 >>>> <http://192.168.1.172:6789/0>" >>>> [node30][DEBUG ] ], >>>> [node30][DEBUG ] "monmap": { >>>> [node30][DEBUG ] "created": "0.000000", >>>> [node30][DEBUG ] "epoch": 0, >>>> [node30][DEBUG ] "fsid": "0d00a742-7ac1-4535-b0dc-26f5a0fe7924", >>>> [node30][DEBUG ] "modified": "0.000000", >>>> [node30][DEBUG ] "mons": [ >>>> [node30][DEBUG ] { >>>> [node30][DEBUG ] "addr": "192.168.1.173:6789/0 >>>> <http://192.168.1.173:6789/0>", >>>> [node30][DEBUG ] "name": "node30", >>>> [node30][DEBUG ] "rank": 0 >>>> [node30][DEBUG ] }, >>>> [node30][DEBUG ] { >>>> [node30][DEBUG ] "addr": "0.0.0.0:0/1 <http://0.0.0.0:0/1>", >>>> [node30][DEBUG ] "name": "node31", >>>> [node30][DEBUG ] "rank": 1 >>>> [node30][DEBUG ] }, >>>> [node30][DEBUG ] { >>>> [node30][DEBUG ] "addr": "0.0.0.0:0/2 <http://0.0.0.0:0/2>", >>>> [node30][DEBUG ] "name": "node32", >>>> [node30][DEBUG ] "rank": 2 >>>> [node30][DEBUG ] } >>>> [node30][DEBUG ] ] >>>> [node30][DEBUG ] }, >>>> [node30][DEBUG ] "name": "node30", >>>> [node30][DEBUG ] "outside_quorum": [ >>>> [node30][DEBUG ] "node30" >>>> [node30][DEBUG ] ], >>>> [node30][DEBUG ] "quorum": [], >>>> [node30][DEBUG ] "rank": 0, >>>> [node30][DEBUG ] "state": "probing", >>>> [node30][DEBUG ] "sync_provider": [] >>>> [node30][DEBUG ] } >>>> >>>> >>>> _______________________________________________ >>>> ceph-users mailing list >>>> ceph-users@xxxxxxxxxxxxxx >>>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>>> >>> >>> >> >> > > > -- > Wido den Hollander > 42on B.V. > > Phone: +31 (0)20 700 9902 > Skype: contact42on > _______________________________________________ > ceph-users mailing list > ceph-users@xxxxxxxxxxxxxx > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |