Hi,
I am running this version of corosync:
corosync-1.2.3-36.el6.x86_64
on a 3-node cluster running RHEL 6 / kernel 2.6.33.9-rt31.75.el6rt.x86_64
I am running pacemaker also, but with no services (yet).
I caused the dump by doing this on the third (last) node:
ifdown ib0 && sleep 5 && ifup ib0.
This is some info from the core file:
...
Core was generated by `corosync'.
Program terminated with signal 6, Aborted.
#0 0x00000032a4a32885 in raise () from /lib64/libc.so.6
Missing separate debuginfos, use: debuginfo-install corosync-1.2.3-36.el6.x86_64
(gdb) where
#0 0x00000032a4a32885 in raise () from /lib64/libc.so.6
#1 0x00000032a4a34065 in abort () from /lib64/libc.so.6
#2 0x00000032a4a2b9fe in __assert_fail_base () from /lib64/libc.so.6
#3 0x00000032a4a2bac0 in __assert_fail () from /lib64/libc.so.6
#4 0x00007f69bbb2eaa6 in ?? () from /usr/lib64/libtotem_pg.so.4
#5 0x00007f69bbb33a93 in ?? () from /usr/lib64/libtotem_pg.so.4
#6 0x00007f69bbb33db9 in ?? () from /usr/lib64/libtotem_pg.so.4
#7 0x00007f69bbb2b2a4 in rrp_deliver_fn () from /usr/lib64/libtotem_pg.so.4
#8 0x00007f69bbb27e4a in ?? () from /usr/lib64/libtotem_pg.so.4
#9 0x00007f69bbb23dba in poll_run () from /usr/lib64/libtotem_pg.so.4
#10 0x0000000000406c1e in main ()
(gdb)
Is this a know bug? How to deal with this? (I am new to corosync - I
just started evaluating it for a product that required H/A - sic! :-)
Also, although it core-dump only once - I can get it in a bind quite
easily bu doing the above (ifdown/ifup).
Very often Totem simply does not recover/does not converge/keeps spinning.
Below is my corosync.conf:
# Please read the corosync.conf.5 manual page
compatibility: none
totem {
version: 2
secauth: off
threads: 0
token: 500
retransmits_before_loss: 5
interface {
ringnumber: 0
bindnetaddr: 10.10.100.0
mcastaddr: 226.94.1.1
mcastport: 5405
ttl: 1
}
}
logging {
fileline: off
to_stderr: no
to_logfile: no
to_syslog: yes
logfile: /var/log/cluster/corosync.log
debug: off
timestamp: on
logger_subsys {
subsys: AMF
debug: off
}
}
amf {
mode: disabled
}
Thanks a lot in advance for any advice on what to do with this problem.
--
Mo Po
_______________________________________________
discuss mailing list
discuss@xxxxxxxxxxxx
http://lists.corosync.org/mailman/listinfo/discuss
