Description of problem: After bringing down the main ethernet interface to test recovery and fencing, stonithd segfaulted on the node with the failed ethernet interface. Version-Release number of selected component: pacemaker-1.1.8-1.el7 Additional info: libreport version: 2.0.12 abrt_version: 2.0.11 backtrace_rating: 4 cmdline: /usr/libexec/pacemaker/stonithd crash_function: strlen kernel: 3.5.0-0.24.el7.x86_64 truncated backtrace: :Thread no. 1 (9 frames) : #0 strlen at ../sysdeps/x86_64/strlen.S : #1 __strdup at strdup.c : #2 create_remote_stonith_op at remote.c : #3 initiate_remote_stonith_op at remote.c : #4 stonith_command at commands.c : #5 st_ipc_dispatch at main.c : #6 _process_request_ at ipcs.c : #7 qb_ipcs_dispatch_connection_request at ipcs.c : #8 gio_read_socket at mainloop.c
Created attachment 608627 [details] File: core_backtrace
Created attachment 608628 [details] File: environ
Created attachment 608629 [details] File: backtrace
Created attachment 608630 [details] File: limits
Created attachment 608631 [details] File: cgroup
Created attachment 608632 [details] File: maps
Created attachment 608633 [details] File: dso_list
Created attachment 608634 [details] File: build_ids
Created attachment 608635 [details] File: var_log_messages
Created attachment 608636 [details] File: open_fds
A related patch has been committed upstream: https://github.com/beekhof/pacemaker/commit/85dfc9c with subject: High: Fencing: Prevent use-of-NULL when fencing is initiated by stonith_admin Further details (if any):