Actions
Bug #50468
openSimultaneous mon daemon crash on cephfs mount
Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:
0%
Source:
Community (user)
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
I have a newly installed 16.2.0 cluster consisting of 5 nodes deployed using cephadm on Ubuntu 18.04.5. I have created three cephfs filesystems:
name: cephfstest1, metadata pool: cephfs.cephfstest1.meta, data pools: [cephfs.cephfstest1.data ] name: cephfsnfs, metadata pool: cephfs.cephfsnfs.meta, data pools: [cephfs.cephfsnfs.data ] name: cephfsnfsv3, metadata pool: cephfs.cephfsnfsv3.meta, data pools: [cephfs.cephfsnfsv3.data ] name: sambatest1, metadata pool: cephfs.sambatest1.meta, data pools: [cephfs.sambatest1.data ]
all created as follows:
ceph fs volume create sambatest1
Mounting the cephfstest1 works either via kernel client or ceph-fuse. Attempting to mount the cephfs "sambatest1" using either method times out and results in all 5 mon daemons simultaneously crashing. All monitors show the same stack trace:
{ "crash_id": "2021-04-20T17:20:50.694919Z_085f320d-276c-4f62-9b60-bc0ba1284813", "timestamp": "2021-04-20T17:20:50.694919Z", "process_name": "ceph-mon", "entity_name": "mon.ceph-01", "ceph_version": "16.2.0", "utsname_hostname": "ceph-01", "utsname_sysname": "Linux", "utsname_release": "4.15.0-141-generic", "utsname_version": "#145-Ubuntu SMP Wed Mar 24 18:08:07 UTC 2021", "utsname_machine": "x86_64", "os_name": "CentOS Linux", "os_id": "centos", "os_version_id": "8", "os_version": "8", "backtrace": [ "/lib64/libpthread.so.0(+0x12b20) [0x7f26f5b68b20]", "gsignal()", "abort()", "/lib64/libstdc++.so.6(+0x9009b) [0x7f26f518609b]", "/lib64/libstdc++.so.6(+0x9653c) [0x7f26f518c53c]", "/lib64/libstdc++.so.6(+0x96597) [0x7f26f518c597]", "/lib64/libstdc++.so.6(+0x967f8) [0x7f26f518c7f8]", "/lib64/libstdc++.so.6(+0x92045) [0x7f26f5188045]", "/usr/bin/ceph-mon(+0x4d78a6) [0x55bfb12c78a6]", "(MDSMonitor::check_sub(Subscription*)+0x819) [0x55bfb12bde69]", "(Monitor::handle_subscribe(boost::intrusive_ptr<MonOpRequest>)+0xcd8) [0x55bfb10b0da8]", "(Monitor::dispatch_op(boost::intrusive_ptr<MonOpRequest>)+0x4cd) [0x55bfb10d6c7d]", "(Monitor::_ms_dispatch(Message*)+0x5f6) [0x55bfb10d8236]", "(Dispatcher::ms_dispatch2(boost::intrusive_ptr<Message> const&)+0x5c) [0x55bfb11067bc]", "(DispatchQueue::entry()+0x126a) [0x7f26f82a7e0a]", "(DispatchQueue::DispatchThread::entry()+0x11) [0x7f26f8357b01]", "/lib64/libpthread.so.0(+0x814a) [0x7f26f5b5e14a]", "clone()" ] }
Attached is the crash log for one of the mon daemons. I also have a capture of the mon log with full debug running which I can supply but it exceeds the attachment limit.
-David
Files
Updated by Sage Weil almost 3 years ago
- Project changed from Ceph to RADOS
- Category deleted (
Monitor)
Actions