Hi, I've got a strange issue when I try to initialize an object repository with ceph-osd --mkfs or when I mount an OSD partition (previously initialized). On Archlinux with a kernel >3.8.5, with Ceph v0.61.2 or v0.62 and XFS as OSD filesystem, each time I run 'ceph-osd -i $i --mkfs' my machine reboot. Same result with vanilla kernel 3.9.2 still on Archlinux, but also in Debian Sid (with ceph v0.61). On a Debian Squeeze + vanilla kernel 3.9.2, no reboot when I run ceph-osd --mkfs or mount partition. With Btrfs or Ext4 everything went fine, same with previous kernel (3.8.5 in Archlinux). If I initialize OSD with kernel 3.8.5, I upgrade kernel and reboot, my host will reboot endless since my OSDs are started at boot. I've reproduced this issue on both real machine and vm, with a newly installed Archlinux and just upgraded Debian from Squeeze to Sid. I don't know if this is linked, but on time I've got a null pointer deference repeated endlessly instead of a reboot: BUG: unable to handle kernel NULL pointer deference at 0000000000000068 IP: [<ffffffff814c8f6b>] no_context+0x211/0x28f PGD 1d35a067 PUD 1d35b067 PMD 0 Any idea of the problem ? Thanks. _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com