The times I have seen this message, it has always been because there are snapshots of the image that haven't been deleted yet. You can see the snapshots with "rbd snap list <image>". On Tue, May 20, 2014 at 4:26 AM, James Eckersall <james.eckersall at gmail.com> wrote: > Hi, > > > > I'm having some trouble with an rbd image. I want to rename the current rbd > and create a new rbd with the same name. > > I renamed the rbd with rbd mv, but it was still mapped on another node, so > rbd mv gave me an error that it was unable to remove the source. > > > I then unmapped the original rbd and tried to remove it. > > > Despite it being unmapped, the cluster still believes that there is a > watcher on the rbd: > > > root at ceph-admin:~# rados -p poolname listwatchers rbdname.rbd > > watcher=x.x.x.x:0/2329830975 client.26367 cookie=48 > > root at ceph-admin:~# rbd rm -p poolname rbdname > > Removing image: 99% complete...failed.2014-05-20 11:50:15.023823 > 7fa6372e4780 -1 librbd: error removing header: (16) Device or resource busy > > > rbd: error: image still has watchers > > This means the image is still open or the client using it crashed. Try again > after closing/unmapping it or waiting 30s for the crashed client to timeout. > > > > I've already rebooted the node that the cluster claims is a watcher and > confirmed it definitely is not mapped. > > I'm 99.9% sure that there are no nodes actually using this rbd. > > > Does anyone know how I can get rid of it? > > > Currently running ceph 0.73-1 on Ubuntu 12.04. > > > Thanks > > > J > > > _______________________________________________ > ceph-users mailing list > ceph-users at lists.ceph.com > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >
 |