Lars Marowsky-Bree wrote:
On 2006-06-24T00:20:12, James Pan <[EMAIL PROTECTED]> wrote:
Hi
Sun Jiang Dong help me on analyzing the result. The cause turned out
to be the crmd which wouldn't exit.
So we sent a signal 15 to the crmd maually( killall -15 crmd) , this
time crmd exited and the testing continued.
This issue happen everytime.
See the attachments for the log, ha.cf, cib.xml.
Could you file a bugzilla for this one instead please?
Ok, bug #1346 :-)
#node halp01 halp02 halp03 halp04 halp05 halp06
node halp01 halp02 halp03 halp04 halp05
ping 9.3.189.1
ucast eth0 9.3.189.146
ucast eth0 9.3.189.147
ucast eth0 9.3.189.148
ucast eth0 9.3.189.149
ucast eth0 9.3.189.150
BTW, this is a rather wasteful network setup. ucast means that the data
will be send to each node independently. You'd use much less bandwidth
with multicast.
Ok, i see, The config file was derived from an exist one, i will change
it to an efficient way
before next run. :-)
auto_failback legacy
Doesn't have any relevance in a crm-style cluster, does it?
Sincerely,
Lars Marowsky-Brée
_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
