On Wed, 30 Jun 2010, Sébastien Paolacci wrote: > On 6/30/10, Sage Weil <sage@xxxxxxxxxxxx> wrote: > > Hmm.. is this reproducible? I just pushed something to ceph.git unstable > > that prints out useful debugging info when that happens, but the question > > remains what is causing it to happen in the first place. Do you mind > > giving the latest a go and seeing what it tells us? > > > Yes it's definitely reproducible. I gave a try to your last commit > (that is actually triggered), here is the log output: Can you do the same, but with debug filestore = 20 debug journal = 20 debug osd = 20 I'm not sure where that length 5120 write is coming from. Thanks! sage > > 10.06.30_22:49:37.214544 --- opened log /var/log/ceph/debian-vm1.11418 --- > ceph version 0.21~rc (3235abe91863b3bffe5638b2169603f7d89ea375) > 10.06.30_22:49:37.214655 7fec9a01c720 ---- renamed symlink > /var/log/ceph/osd0 -> /var/log/ceph/osd0.0 ---- > 10.06.30_22:49:37.214689 7fec9a01c720 ---- created symlink > /var/log/ceph/osd0 -> debian-vm1.11418 ---- > 10.06.30_22:49:37.214747 7fec9a01c720 -- :/0 register_entity client? > 10.06.30_22:49:37.214876 7fec9a01c720 -- :/0 register_entity client? at :/0 > 10.06.30_22:49:37.214953 7fec9a01c720 -- :/0 ready :/0 > 10.06.30_22:49:37.215019 7fec9a01c720 -- :/11418 messenger.start > 10.06.30_22:49:37.215076 7fec9a01c720 -- :/11418 shutdown :/11418 > 10.06.30_22:49:37.215091 7fec9a01c720 -- :/11418 shutdown i am not > dispatch, setting stop flag and joining thread. > 10.06.30_22:49:37.215107 7fec9a01c720 -- :/11418 wait: still active > 10.06.30_22:49:37.215145 7fec9a01c720 -- :/11418 wait: woke up > 10.06.30_22:49:37.215165 7fec9a01c720 -- :/11418 wait: everything stopped > 10.06.30_22:49:37.215177 7fec9a01c720 -- :/11418 wait: closing pipes > 10.06.30_22:49:37.215191 7fec9a01c720 -- :/11418 reaper > 10.06.30_22:49:37.215202 7fec9a01c720 -- :/11418 reaper done > 10.06.30_22:49:37.215214 7fec9a01c720 -- :/11418 wait: waiting for > pipes to close > 10.06.30_22:49:37.215225 7fec9a01c720 -- :/11418 wait: done. > 10.06.30_22:49:37.215236 7fec9a01c720 -- :/11418 shutdown complete. > 10.06.30_22:49:37.215398 7fec9a01c720 filestore(/data/osd0) mkfs in /data/osd0 > 10.06.30_22:49:37.215477 7fec9a01c720 filestore(/data/osd0) mkfs > removing old directory current > 10.06.30_22:49:37.218588 7fec9a01c720 filestore(/data/osd0) mkfs > removing old file magic > 10.06.30_22:49:37.218644 7fec9a01c720 filestore(/data/osd0) mkfs > removing old file fsid > 10.06.30_22:49:37.218674 7fec9a01c720 filestore(/data/osd0) mkfs > removing old file ceph_fsid > 10.06.30_22:49:37.218703 7fec9a01c720 filestore(/data/osd0) mkfs > removing old file whoami > 10.06.30_22:49:37.219248 7fec9a01c720 filestore(/data/osd0) mkjournal > created journal on /ceph-journal/osd0 > 10.06.30_22:49:37.219279 7fec9a01c720 filestore(/data/osd0) mkfs done > in /data/osd0 > 10.06.30_22:49:37.219329 7fec9a01c720 filestore(/data/osd0) mount did > NOT detect btrfs > 10.06.30_22:49:37.219376 7fec9a01c720 filestore(/data/osd0) mount found snaps <> > 10.06.30_22:49:37.220478 7fec98932710 journal rebuild_page_aligned > failed, buffer::list(len=5120, > buffer::ptr(0~5120 0xab3410 in raw 0xab3410 len 5120 nref 1) > ) > os/FileJournal.cc: In function 'void FileJournal::write_bl(off64_t&, > ceph::bufferlist&)': > os/FileJournal.cc:508: FAILED assert((bl.length() & ~ceph::_page_mask) == 0) > 1: (FileJournal::do_write(ceph::buffer::list&)+0x2a2) [0x55b362] > 2: (FileJournal::write_thread_entry()+0x1fa) [0x55da6a] > 3: (FileJournal::Writer::entry()+0xd) [0x54f04d] > 4: (Thread::_entry_func(void*)+0x7) [0x46ca97] > 5: (()+0x68ba) [0x7fec999fe8ba] > 6: (clone()+0x6d) [0x7fec98c1901d] > NOTE: a copy of the executable, or `objdump -rdS <executable>` is > needed to interpret this. > > Would you prefer to have the core dump? > > Sebastien > -- > 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 > >
