Yes, you've run across an issue. It's identified and both the preventive fix and the resolver tool are in testing. See: The thread "Emperor upgrade bug 6761" http://tracker.ceph.com/issues/6761#note-19 (Forgive my brevity; it's late here. :) -Greg Software Engineer #42 @ http://inktank.com | http://ceph.com On Thu, Nov 14, 2013 at 2:15 AM, Damien Churchill <damoxc@xxxxxxxxx> wrote: > Hi, > > Since the upgrade to 0.72 I've been experiencing an issue with a number of > volumes. It seems to occur on both the librbd clients as well as the krbd > clients. From the kernel client, when trying to access one of these > appearing corrupted images, it causes a panic and you end up with an > assertion failure in syslog. > > [1483456.028548] rbd: rbd9: read a000 at ed000000 (0) > [1483456.028548] > [1483456.028553] rbd: rbd9: result -23 xferred 0 > [1483456.028553] > [1483456.028556] end_request: I/O error, dev rbd9, sector 7766016 > [1483456.028889] > [1483456.028889] Assertion failure in rbd_img_obj_callback() at line 2141: > [1483456.028889] > [1483456.028889] rbd_assert(more ^ (which == > img_request->obj_request_count)); > > Looking at the code for rbd_img_obj_callback() it seems that the wrong > number of objects are being returned. > > When using userland tools they exit with errno 23 which is a little > confusing as that maps to "Too many open files in system". It's possible to > perform a rbd copy but that's the only operation that is successful. > > Has anyone experienced this before, or able to give any pointers about what > is going wrong here? > > Thanks, > Damien > > _______________________________________________ > 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
 |