Dear Ceph Team, Our company was trying to using ceph as our backend support. And I’ve encountered an error while deploying multiple clusters on the same servers. I wonder whether this is a bug or I did somthing wrong. According to the documents on ceph.com, I successfully built a ceph cluster contains 2 mons, 2 osds & 1 mds on 2 servers via ceph-deploy & (installation quick tutorial on ceph.com). I’ve noticed that the ceph supports deploying multiple cluster on the same hardware. So I try to use ceph-deploy on the same servers to create another cluster and I did execute all the ceph-deploy commands argumented with "—cluster ceph2” for a new cluster named “ceph2”. Below are command sequence I’ve executed: # dev208 & dev209 already built a ceph cluster named with default name “ceph” # now I’m trying to create a new cluster named ceph2 mkdir cluster2 && cd cluster2 ceph-deploy —cluster ceph2 new dev208 # dev208 is the server hostname ceph-deploy —cluster ceph2 mon create-initial # an error occurred in this step … Below are some of logs: dev208][DEBUG ] determining if provided host has same hostname in remote [dev208][DEBUG ] get remote short hostname [dev208][DEBUG ] deploying mon to dev208 [dev208][DEBUG ] get remote short hostname [dev208][DEBUG ] remote hostname: dev208 [dev208][DEBUG ] write cluster configuration to /etc/ceph/{cluster}.conf [dev208][DEBUG ] create the mon path if it does not exist [dev208][DEBUG ] checking for done path: /var/lib/ceph/mon/ceph2-dev208/done [dev208][DEBUG ] done path does not exist: /var/lib/ceph/mon/ceph2-dev208/done [dev208][INFO ] creating keyring file: /var/lib/ceph/tmp/ceph2-dev208.mon.keyring [dev208][DEBUG ] create the monitor keyring file [dev208][INFO ] Running command: sudo ceph-mon --cluster ceph2 --mkfs -i dev208 --keyring /var/lib/ceph/tmp/ceph2-dev208.mon.keyring [dev208][DEBUG ] ceph-mon: renaming mon.noname-a 192.168.0.208:6789/0 to mon.dev208 [dev208][DEBUG ] ceph-mon: set fsid to 0df0dff1-5a6d-431f-9a28-4ef8fd53c207 [dev208][DEBUG ] ceph-mon: created monfs at /var/lib/ceph/mon/ceph2-dev208 for mon.dev208 [dev208][INFO ] unlinking keyring file /var/lib/ceph/tmp/ceph2-dev208.mon.keyring [dev208][DEBUG ] create a done file to avoid re-doing the mon deployment [dev208][DEBUG ] create the init path if it does not exist [dev208][DEBUG ] locating the `service` executable... [dev208][INFO ] Running command: sudo /usr/sbin/service ceph -c /etc/ceph/ceph2.conf start mon.dev208 [dev208][DEBUG ] === mon.dev208 === [dev208][DEBUG ] Starting Ceph mon.dev208 on dev208...already running [dev208][INFO ] Running command: sudo systemctl enable ceph [dev208][WARNING] ceph.service is not a native service, redirecting to /sbin/chkconfig. [dev208][WARNING] Executing /sbin/chkconfig ceph on [dev208][WARNING] The unit files have no [Install] section. They are not meant to be enabled [dev208][WARNING] using systemctl. [dev208][WARNING] Possible reasons for having this kind of units are: [dev208][WARNING] 1) A unit may be statically enabled by being symlinked from another unit's [dev208][WARNING] .wants/ or .requires/ directory. [dev208][WARNING] 2) A unit's purpose may be to act as a helper for some other unit which has [dev208][WARNING] a requirement dependency on it. [dev208][INFO ] Running command: sudo ceph --cluster=ceph2 --admin-daemon /var/run/ceph/ceph2-mon.dev208.asok mon_status [dev208][ERROR ] admin_socket: exception getting command descriptions: [Errno 2] No such file or directory I’ve noticed that the mds for new cluster ceph2 is not started since there are already a ceph monitor running. so it cannot find the /var/run/ceph/ceph2-mon.dev208.asok. It seems that the service script /etc/init.d/ceph does not handle correctly about multiple ceph services on the same server. Could you give some guidances on deploying multiple clusters on the same servers? Best regards, |
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
