Hi Sandor, the deploy script is in the first place a script to maintain existing ceph environments and build up new. If you want to use existing keys/informations you will have to modify the deploy script according to your needs. -- What should work would be, if you create your new cluster with the same Cluster ID ( fsid ) without using old keyrings/what ever. Then you copy everything ( /var/lib/ceph/* and /etc/ceph/* and everything i forgot ^^; ) and reuse your OSD's. If your new cluster has a new fsid, you might run very fast into trouble. If your old cluster has more/less nodes/monitors/osd's/... then this way would also not work properly i think. So in the very end, the most secure way would be to create two independent clusters and simply copy the data. Using your old OSD's will require your old informations ( which are based on the specific setup your old cluster has ( number of nodes/monitors/osd's/size/... ). If you change there something, you are automatically in the experimental/danger zone :-) Good luck ! -- Mit freundlichen Gruessen / Best regards Oliver Dzombic IP-Interactive mailto:info@xxxxxxxxxxxxxxxxx Anschrift: IP Interactive UG ( haftungsbeschraenkt ) Zum Sonnenberg 1-3 63571 Gelnhausen HRB 93402 beim Amtsgericht Hanau Geschäftsführung: Oliver Dzombic Steuer Nr.: 35 236 3622 1 UST ID: DE274086107 Am 07.03.2016 um 14:38 schrieb Sándor Szombat: > Hi guys! > > I want to port my cluster without using rados import/export function. > The scenario is next: I made a cluster with 3 node, I fill it with data. > After this I run /ceph-deploy gatherkeys host1 /and /ceph-deploy config > pull host1/ -> I have keyring files and config file. I make an image > from OSD hdds. > Then I take 3 new node. I run ceph-deploy commands with --keyrings > parameter (using keyrings and config from three old node). After this > I've got same cluster? Can I reuse OSDs? I tried it but unfortunelly I > can't run ceph -s because it can't find */ceph.client.admin.keyring > file/*. But my question: can we port ceph cluster with ceph-deploy tool > or not? > > Sry for my bad english and thanks for your help :) > > > > > _______________________________________________ > 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
 |