[Gluster-devel] gluster server 3.0 segfaults on ibverbs setup

[Anders Litzell]

Hi everyone,

What I did:
I did an upgrade to an existing very small lab setup (one server, one 
brick, localhost only client). yum erase old rpm, yum localinstall new 
rpm and no change to config files, added --debug to init script.

The result:
A very similar segfault error to the original post and I'm hoping my 
data may be able to shed more light on this. Please see debug output 
below. Segfault happens 2-10 seconds after daemon start.

As a curiosity please note that nothing is written to the log files. All 
output to stdout EXCEPT last line with the segfault which is to stderr.

[root at wagner64 mnt]# /etc/init.d/glusterfsd start
[2009-12-10 15:39:51] D [glusterfsd.c:424:_get_specfp] glusterfs: 
loading volume file /etc/glusterfs/glusterfsd.vol
[2009-12-10 15:39:51] D [xlator.c:739:xlator_set_type] xlator: 
dlsym(notify) on /usr/lib64/glusterfs/3.0.0/xlator/features/locks.so: 
undefined symbol: notify -- neglecting
[2009-12-10 15:39:51] D [xlator.c:739:xlator_set_type] xlator: 
dlsym(notify) on 
/usr/lib64/glusterfs/3.0.0/xlator/performance/io-threads.so: undefined 
symbol: notify -- neglecting
[2009-12-10 15:39:51] D [xlator.c:744:xlator_set_type] xlator: 
dlsym(dumpops) on 
/usr/lib64/glusterfs/3.0.0/xlator/performance/io-threads.so: undefined 
symbol: dumpops -- neglecting
================================================================================
Version      : glusterfs 3.0.0 built on Dec  8 2009 03:09:12
git: 2.0.1-886-g8379edd
Starting Time: 2009-12-10 15:39:51
Command line : /usr/sbin/glusterfsd -f /etc/glusterfs/glusterfsd.vol 
--debug
PID          : 574
System name  : Linux
Nodename     : wagner64
Kernel Release : 2.6.18-164.6.1.el5
Hardware Identifier: x86_64

Given volfile:
+------------------------------------------------------------------------------+
  1: volume posix0
  2:   type storage/posix
  3:   option directory /mnt/aoe_90/export
  4: end-volume
  5:
  6: volume locks0
  7:     type features/locks
  8:     subvolumes posix0
  9: end-volume
 10:
 11: volume brick0
 12:     type performance/io-threads
 13:     option thread-count 8
 14:     subvolumes locks0
 15: end-volume
 16:
 17: volume server
 18:     type protocol/server
 19:     option transport-type tcp
 20:     option auth.addr.brick0.allow *
 21:     option listen-port 6996
 22:     subvolumes brick0
 23: end-volume

+------------------------------------------------------------------------------+
[2009-12-10 15:39:51] D [glusterfsd.c:1335:main] glusterfs: running in 
pid 574
[2009-12-10 15:39:51] D [transport.c:145:transport_load] transport: 
attempt to load file /usr/lib64/glusterfs/3.0.0/transport/socket.so
[2009-12-10 15:39:51] W [xlator.c:655:validate_xlator_volume_options] 
server: option 'listen-port' is deprecated, preferred is 
'transport.socket.listen-port', continuing with correction
[2009-12-10 15:39:51] D [xlator.c:284:_volume_option_value_validate] 
server: no range check required for 'option transport.socket.listen-port 
6996'
[2009-12-10 15:39:51] D [io-threads.c:2841:init] brick0: io-threads: 
Autoscaling: off, min_threads: 8, max_threads: 8
[2009-12-10 15:39:51] N [glusterfsd.c:1361:main] glusterfs: Successfully 
started
pending frames:
/bin/bash: line 1:   574 Segmentation fault      (core dumped) 
/usr/sbin/glusterfsd -f /etc/glusterfs/glusterfsd.vol --debug
-------- end of pasted output ---------

Hope my data helps.

Cheers

Anders