Per http://tracker.ceph.com/issues/6022 leveldb-1.12 was pulled out of
the ceph-extras repo due to patches applied by a leveldb fork (Basho
patch). It's back in ceph-extras (since the 28th at least), and on
CentOS 6 is causing an abort on mon start when run with the Firefly
release candidate
# /etc/init.d/ceph start
=== mon.m1 ===
Starting Ceph mon.m1 on m1...
pthread lock: Invalid argument
*** Caught signal (Aborted) **
in thread 7fa7524f67a0
ceph version 0.80-rc1-28-ga027100
(a0271000c12486d3c5adb2b0732e1c70c3789a4f)
1: /usr/bin/ceph-mon() [0x86bcb1]
2: /lib64/libpthread.so.0() [0x35cac0f710]
3: (gsignal()+0x35) [0x35ca432925]
4: (abort()+0x175) [0x35ca434105]
5: (()+0x34d71) [0x7fa752532d71]
6: (leveldb::DBImpl::Get(leveldb::ReadOptions const&, leveldb::Slice
const&, leveldb::Value*)+0x50) [0x7fa7
52518120]
7: (LevelDBStore::_get_iterator()+0x41) [0x826d71]
8: (MonitorDBStore::exists(std::string const&, std::string
const&)+0x28) [0x539e88]
9: (main()+0x13f8) [0x533d78]
10: (__libc_start_main()+0xfd) [0x35ca41ed1d]
11: /usr/bin/ceph-mon() [0x530f19]
2014-04-30 10:32:40.397243 7fa7524f67a0 -1 *** Caught signal (Aborted) **
in thread 7fa7524f67a0
The SRPM for what ended up on ceph-extras wasn't uploaded to the repo,
so I didn't check to see if it was the Basho patch being applied again
or something else. Downgrading back to leveldb 1.7.0-2 resolved my problem.
Is anyone else seeing this?
Jeff
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
