Hi Josh, i quoted the trace and some other stats in my first email, maybe it got stuck in the spam filters. Well next try: snip -3> 2012-05-10 14:52:29.509940 7fb1c9351700 1 mds.0.40 handle_mds_map i am now mds.0.40 -2> 2012-05-10 14:52:29.509956 7fb1c9351700 1 mds.0.40 handle_mds_map state change up:reconnect --> up:rejoin -1> 2012-05-10 14:52:29.509963 7fb1c9351700 1 mds.0.40 rejoin_joint_start 0> 2012-05-10 14:52:29.512503 7fb1c9351700 -1 *** Caught signal (Segmentation fault) ** in thread 7fb1c9351700 ceph version 0.46 (commit:cb7f1c9c7520848b0899b26440ac34a8acea58d1) 1: ceph-mds() [0x814279] 2: (()+0xeff0) [0x7fb1cddbfff0] 3: (SnapRealm::have_past_parents_open(snapid_t, snapid_t)+0x4f) [0x6cb5ef] 4: (MDCache::check_realm_past_parents(SnapRealm*)+0x2b) [0x55d58b] 5: (MDCache::choose_lock_states_and_reconnect_caps()+0x29c) [0x572eec] 6: (MDCache::rejoin_gather_finish()+0x90) [0x5931a0] 7: (MDCache::rejoin_send_rejoins()+0x2c05) [0x59b9d5] 8: (MDS::rejoin_joint_start()+0x131) [0x4a8721] 9: (MDS::handle_mds_map(MMDSMap*)+0x2c4a) [0x4c253a] 10: (MDS::handle_core_message(Message*)+0x913) [0x4c4513] 11: (MDS::_dispatch(Message*)+0x2f) [0x4c45ef] 12: (MDS::ms_dispatch(Message*)+0x1fb) [0x4c628b] 13: (SimpleMessenger::dispatch_entry()+0x979) [0x7acb49] 14: (SimpleMessenger::DispatchThread::entry()+0xd) [0x7336ed] 15: (()+0x68ca) [0x7fb1cddb78ca] 16: (clone()+0x6d) [0x7fb1cc63f92d] snip I though ceph chooses which MDS is active and which is standby, i just have 3 in the cluster config: [mds.a] host = x [mds.b] host = y [mds.c] host = z no global MDS config. Should i reconfigure this? 2012/5/17 Josh Durgin <josh.durgin@xxxxxxxxxxx>: > On 05/16/2012 01:11 AM, Felix Feinhals wrote: >> >> Hi again, >> >> anything on this Problem? Seems that the only choice for me is to >> reinitialize the whole cephfs (mkcephfs...) >> :( > > > Hi Felix, it looks like your first mail never reached the list. > > >> 2012/5/10 Felix Feinhals<ff@xxxxxxxxxxxxxxxxxxxxxxx>: >>> >>> Hi List, >>> >>> we installed a ceph cluster with ceph version 0.46. >>> 3 OSDs, 3 MONs and 3 MDSs. >>> >>> After copying a bunch of files to a ceph-fuse mount all MDS daemons >>> crash and now i cant bring them back online. >>> I already tried to restart the daemons in different order and also >>> removed one OSD, nothing really happened only now we have pgs with >>> active+remapped which i think is normal. >>> Any hints? > > > Are all three MDS active? At this point, more than one active MDS is > likely to crash. You can have one active and others standby. > > If you've got only one active, what was the backtrace of the crash? > It'll be at the end of the MDS log (by default in /var/log/ceph). -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
