Hello,
my ceph-osd luminous are crashing with a segmentation fault while
backfilling.
Is there any way to manually remove the problematic "data"?
-1> 2018-01-16 20:32:50.001722 7f27d53fe700 0 osd.86 pg_epoch:
917877 pg[3.80e( v 917875'69934125 (917365'69924082,917875'69934125] lb
3:7018abae:::rbd_data.1ba91116b8b4567.0000000000004362:head (bitwise)
local-lis/les=913221/913222 n=895 ec=15/15 lis/c 913221/909473 les/c/f
913222/909474/0 917852/917852/917219) [50,54,86]/[54] r=-1 lpr=917852
pi=[909473,917852)/11 luod=0'0 crt=917875'69934125 lcod 917875'69934124
active+remapped] snapset b0cee=[b0cee]:{} legacy_snaps []
0> 2018-01-16 20:32:50.004728 7f27d53fe700 -1 *** Caught signal
(Segmentation fault) **
in thread 7f27d53fe700 thread_name:tp_osd_tp
ceph version 12.2.2-93-gd6da8d7
(d6da8d77a4b2220e6bdd61e4bdd911a9cd91946c) luminous (stable)
1: (()+0xa43dec) [0x563de6597dec]
2: (()+0xf890) [0x7f282f7fc890]
3: (std::_Rb_tree_iterator<snapid_t> std::_Rb_tree<snapid_t, snapid_t,
std::_Identity<snapid_t>, std::less<snapid_t>, std::allocator<snapid_t>
>::_M_insert_unique_<snapid_t&>(std::_Rb_tree_const_iterator<snapid_t>,
snapid_t&)+0x40) [0x563de612f6c0]
4: (PrimaryLogPG::on_local_recover(hobject_t const&, ObjectRecoveryInfo
const&, std::shared_ptr<ObjectContext>, bool,
ObjectStore::Transaction*)+0xaae) [0x563de6184fee]
5: (ReplicatedBackend::handle_push(pg_shard_t, PushOp const&,
PushReplyOp*, ObjectStore::Transaction*)+0x31d) [0x563de62f71dd]
6: (ReplicatedBackend::_do_push(boost::intrusive_ptr<OpRequest>)+0x18f)
[0x563de62f747f]
7:
(ReplicatedBackend::_handle_message(boost::intrusive_ptr<OpRequest>)+0x2d1)
[0x563de6307521]
8: (PGBackend::handle_message(boost::intrusive_ptr<OpRequest>)+0x50)
[0x563de622ce40]
9: (PrimaryLogPG::do_request(boost::intrusive_ptr<OpRequest>&,
ThreadPool::TPHandle&)+0x77b) [0x563de619914b]
10: (OSD::dequeue_op(boost::intrusive_ptr<PG>,
boost::intrusive_ptr<OpRequest>, ThreadPool::TPHandle&)+0x3f7)
[0x563de6025bc7]
11: (PGQueueable::RunVis::operator()(boost::intrusive_ptr<OpRequest>
const&)+0x57) [0x563de629d947]
12: (OSD::ShardedOpWQ::_process(unsigned int,
ceph::heartbeat_handle_d*)+0x108c) [0x563de6054d1c]
13: (ShardedThreadPool::shardedthreadpool_worker(unsigned int)+0x88d)
[0x563de65e0e6d]
14: (ShardedThreadPool::WorkThreadSharded::entry()+0x10) [0x563de65e2e30]
15: (()+0x8064) [0x7f282f7f5064]
16: (clone()+0x6d) [0x7f282e8e962d]
NOTE: a copy of the executable, or `objdump -rdS <executable>` is
needed to interpret this.
Greets,
Stefan
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com