[Linux-ha-dev] stonithd received sigabt when testing by CTS.

[AOSHIMA Kentaro]

Hi,all

I'm Aoshima. I'd like to ask about CTS for 2.1.4.
We try to run CTS on RHEL5.2(x86-64) for 2.1.4 whose
changeset is "12148:e902ad7642fd" .

heartbeat and resource agent seemed to fail to start
after kicking CTSlab.py .
According to ha-log, when the lrmd procces tried to start
STONITH resource, something wrong had occured and  stonithd had
received SIGABT.

When using stonith(ex: external/ssh) for CTS,
any extra configurations are needed?

best regards

our testing environment are indicated below.
---ha.cf---
crm on
logfile  /var/log/ha-log
logfacility local7
keepalive 2
deadtime 30
warntime 10
initdead 60
udpport 651
auto_failback legacy
bcast eth1
bcast eth3

node cgl49
node cgl50

--ha-log--
lrmd[8899]: 2008/07/22_14:12:04 info: rsc:ocf_msdummy:0: start
lrmd[9361]: 2008/07/22_14:12:04 info: Try to start STONITH resource
<rsc_id=child_DoFencing:0> : Device=external/ssh
stonithd[8900]: 2008/07/22_14:12:04 ERROR: ipc_bufpool_update: magic
number in head does not match.Something very bad happened, abor
t now, farside pid =9361
stonithd[8900]: 2008/07/22_14:12:04 ERROR: magic=6c754a3e, expected
value=abcd
stonithd[8900]: 2008/07/22_14:12:04 info: pool: refcount=1,
startpos=0x17dc4ba8, currpos=0x17dc4c8d,consumepos=0x17dc4c14, endpos=0x
17dc5b78, size=4096
stonithd[8900]: 2008/07/22_14:12:04 info: nmsgs=0
lrmd[8899]: 2008/07/22_14:12:04 info: rsc:ocf_msdummy:1: start
lrmd[9361]: 2008/07/22_14:12:04 ERROR: stonithd_virtual_stonithRA_ops:
failed to fetch reply
tengine[8951]: 2008/07/22_14:12:04 ERROR: stonithd_op_result_ready: not
signed on
heartbeat[8887]: 2008/07/22_14:12:04 WARN: Managed
/usr/lib64/heartbeat/stonithd process 8900 killed by signal 6 [SIGABRT -
Abort].
lrmd[9361]: 2008/07/22_14:12:04 ERROR: sending stonithRA op to stonithd
failed.
tengine[8951]: 2008/07/22_14:12:04 ERROR:
tengine_stonith_connection_destroy: Fencing daemon has left us
heartbeat[8887]: 2008/07/22_14:12:04 ERROR: Managed
/usr/lib64/heartbeat/stonithd process 8900 dumped core
lrmd[9361]: 2008/07/22_14:12:04 notice: Not currently connected.
tengine[8951]: 2008/07/22_14:12:04 info: te_connect_stonith: Attempting
connection to fencing daemon...
heartbeat[8887]: 2008/07/22_14:12:04 ERROR: Respawning client
"/usr/lib64/heartbeat/stonithd":
heartbeat[8887]: 2008/07/22_14:12:04 info: Starting child client
"/usr/lib64/heartbeat/stonithd" (0,0)
lrmd[8899]: 2008/07/22_14:12:04 WARN: mapped the invalid return code 254.
crmd[8902]: 2008/07/22_14:12:04 info: process_lrm_event: LRM operation
child_DoFencing:0_start_0 (call=19, rc=1) complete
heartbeat[9368]: 2008/07/22_14:12:04 info: Starting
"/usr/lib64/heartbeat/stonithd" as uid 0  gid 0 (pid 9368)