On Tue, 9 Nov 2010, Theodore Ts'o wrote: > > Hi there. I tried building the latest in the testing branch (commit id > e90a3b623), and things didn't go all that well. Using a 2.6.36 client > with commit d91f2438 reverted and the "ceph: fix small seq message > skipping" comimt added, I was able to mount the test file system, but > when I tried to list a directory in the ceph file system, the ls command > hung forever. Dmesg show that it was a problem communicating with > mds0. When I looked on that machine's log, I found the attached stack > dump. Was it a fresh file system, or something with data from before? > I'm going to back off to 0.22.2, which worked much better for me, but I > figured you'd probably be interested in the heads up. > > Regards, > > - Ted > > mds/CInode.cc: In function 'void CInode::finish_scatter_gather_update(int)': > mds/CInode.cc:1479: FAILED assert(pi->dirstat.size() >= 0) I don't think think this code was touched in that branch since 0.22.2. However, it was heavily revised for 0.23, and all those changes are in the 'rc' branch. I think you'll have better luck there. sage > ceph version 0.22.2 (commit:652417c19f5d18c70d5928e67f72c3770c3c57c5) > 1: (CInode::finish_scatter_gather_update(int)+0x839) [0x8ee20d] > 2: (Locker::scatter_writebehind(ScatterLock*)+0x237) [0x88b2f1] > 3: (Locker::simple_sync(SimpleLock*, bool*)+0x350) [0x88a44e] > 4: (Locker::_rdlock_kick(SimpleLock*)+0xc2) [0x87e30c] > 5: (Locker::rdlock_start(SimpleLock*, MDRequest*, bool)+0x1dc) [0x87e778] > 6: (Locker::acquire_locks(MDRequest*, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&)+0x1b3d) [0x87be7f] > 7: (Server::handle_client_stat(MDRequest*)+0x33b) [0x76d2a7] > 8: (Server::dispatch_client_request(MDRequest*)+0x2a5) [0x767caf] > 9: (Server::handle_client_request(MClientRequest*)+0x938) [0x7679f4] > 10: (Server::dispatch(Message*)+0x358) [0x760680] > 11: (MDS::_dispatch(Message*)+0x137a) [0x72c97e] > 12: (MDS::ms_dispatch(Message*)+0x38) [0x72b48a] > 13: (Messenger::ms_deliver_dispatch(Message*)+0x63) [0x716a2d] > 14: (SimpleMessenger::dispatch_entry()+0x5d4) [0x707d46] > 15: (SimpleMessenger::DispatchThread::entry()+0x2c) [0x6fcd40] > 16: (Thread::_entry_func(void*)+0x23) [0x715935] > 17: (()+0x6b9a) [0x2b1c972ccb9a] > 18: (clone()+0x6d) [0x2b1c97cb253d] > NOTE: a copy of the executable, or `objdump -rdS <executable>` is needed to interpret this. > *** Caught signal (ABRT) *** > ceph version 0.22.2 (commit:652417c19f5d18c70d5928e67f72c3770c3c57c5) > 1: (ceph::BackTrace::BackTrace(int)+0x2d) [0x9d8e93] > 2: (sigabrt_handler(int)+0x49) [0x9e8ffd] > 3: (()+0x1e9e0) [0x2b1c97cab9e0] > 4: (gsignal()+0x35) [0x2b1c97cc89f5] > 5: (abort()+0x183) [0x2b1c97ccc413] > 6: (__gnu_cxx::__verbose_terminate_handler()+0x115) [0x2b1c975a5275] > 7: (()+0xbf696) [0x2b1c975a3696] > 8: (()+0xbf6c3) [0x2b1c975a36c3] > 9: (()+0xbf7c6) [0x2b1c975a37c6] > 10: (ceph::__ceph_assert_fail(char const*, char const*, int, char const*)+0x24f) [0x9d8ca4] > 11: (CInode::finish_scatter_gather_update(int)+0x839) [0x8ee20d] > 12: (Locker::scatter_writebehind(ScatterLock*)+0x237) [0x88b2f1] > 13: (Locker::simple_sync(SimpleLock*, bool*)+0x350) [0x88a44e] > 14: (Locker::_rdlock_kick(SimpleLock*)+0xc2) [0x87e30c] > 15: (Locker::rdlock_start(SimpleLock*, MDRequest*, bool)+0x1dc) [0x87e778] > 16: (Locker::acquire_locks(MDRequest*, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&, std::set<SimpleLock*, std::less<SimpleLock*>, std::allocator<SimpleLock*> >&)+0x1b3d) [0x87be7f] > 17: (Server::handle_client_stat(MDRequest*)+0x33b) [0x76d2a7] > 18: (Server::dispatch_client_request(MDRequest*)+0x2a5) [0x767caf] > 19: (Server::handle_client_request(MClientRequest*)+0x938) [0x7679f4] > 20: (Server::dispatch(Message*)+0x358) [0x760680] > 21: (MDS::_dispatch(Message*)+0x137a) [0x72c97e] > 22: (MDS::ms_dispatch(Message*)+0x38) [0x72b48a] > 23: (Messenger::ms_deliver_dispatch(Message*)+0x63) [0x716a2d] > 24: (SimpleMessenger::dispatch_entry()+0x5d4) [0x707d46] > 25: (SimpleMessenger::DispatchThread::entry()+0x2c) [0x6fcd40] > 26: (Thread::_entry_func(void*)+0x23) [0x715935] > 27: (()+0x6b9a) [0x2b1c972ccb9a] > 28: (clone()+0x6d) [0x2b1c97cb253d] > -- > 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 > > -- 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