Ah, I'm not totally sure but I think, it worked for me when I did the "new" task with all 3 nodes and then I deployed first one mon and gathered the keys from there and deployed then the other two. Try this perhaps. Greetings Am 24.05.2013 19:17, schrieb Abel Lopez: > Thanks for the suggestion. I tried adding shortnames to /etc/hosts (I > already have DNS setup), but gatherkeys still fails. > > I really want to use ceph for my cinder backend, and also for my > glance, but if I can't deploy it, I won't use it. > > > On Fri, May 24, 2013 at 10:06 AM, Roman Dilken <rdilken@xxxxxx > <mailto:rdilken@xxxxxx>> wrote: > > Hello, > > I'm also new and I played yesterday with ceph-deploy. > > The error ocurred also to me which I could resolve putting all short > servernames with their IPs in /etc/hosts on all nodes. > > A second error occured deploying OSDs with ceph-deploy osd... > I resolved this in three steps: > > 1. I used ceph-deploy osd prepare /dev/xvdb > 2. I mounted the prepared disk and corrected the symlink journal which > pointed to a non-reachable path to /dev/xvdb2 on every osd-node > 3. ceph-deplay osd acivate /dev/xvdb1 worked finally (the partition is > necessary becaus cheph-deploy creates two partitions: 1 for the > osd and > 1 for the journal. > > Greetings, > > Roman > > > > > Am 24.05.2013 18:56, schrieb Abel Lopez: > > Hello all, > > New to the list. > > > > I'm trying to form a repeatable deployment method for our > environments. > > > > Using Ubuntu 12.04, and trying to install Cuttlefish. > > We have an existing chef setup for deploying openstack, so the > > ceph-cookbook sounded like an obvious choice. After following the > > instructions in the readme and the wiki, I see errors when > chef-client runs > > > > [2013-05-24T16:41:31+00:00] INFO: Processing ruby_block[get > > osd-bootstrap keyring] action create (ceph::mon line 96) > > > > When I take a look at the command in this block, I get these errors > > > > ceph auth get-key client.bootstrap-osd > > 2013-05-24 16:42:10.325822 7f0f8e8e6700 0 -- :/17103 >> > > 192.168.112.82:6789/0 <http://192.168.112.82:6789/0> > <http://192.168.112.82:6789/0> pipe(0x2fe7640 sd=3 > > :0 s=1 pgs=0 cs=0 l=1).fault > > 2013-05-24 16:42:16.326438 7f0f8e7e5700 0 -- > 192.168.112.82:0/17103 <http://192.168.112.82:0/17103> > > <http://192.168.112.82:0/17103> >> 192.168.112.82:6789/0 > <http://192.168.112.82:6789/0> > > <http://192.168.112.82:6789/0> pipe(0x7f0f84001d50 sd=3 :0 s=1 pgs=0 > > cs=0 l=1).fault > > > > So, I tried ceph-deploy, which gives me a whole different set of > errors > > ceph-deploy new is fine, mon create is fine, but gatherkeys fails > > ceph-deploy gatherkeys s7r2 s8r2 s9r2 > > Unable to find /etc/ceph/ceph.client.admin.keyring on ['s7r2', > 's8r2', > > 's9r2'] > > Unable to find /var/lib/ceph/bootstrap-osd/ceph.keyring on ['s7r2', > > 's8r2', 's9r2'] > > Unable to find /var/lib/ceph/bootstrap-mds/ceph.keyring on ['s7r2', > > 's8r2', 's9r2'] > > > > I googled around and found this > > http://tracker.ceph.com/issues/4924 > > It says it's resolved, but I couldn't find the branch mentioned > in github. > > > > As a new user, trying to build production environments reliably, I'm > > seeing a lot of barriers. > > What other methods do people use? > > > > > > > > _______________________________________________ > > ceph-users mailing list > > ceph-users@xxxxxxxxxxxxxx <mailto:ceph-users@xxxxxxxxxxxxxx> > > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > > > > _______________________________________________ > ceph-users mailing list > ceph-users@xxxxxxxxxxxxxx <mailto: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
 |