I setup 2 jboss on two machines.
Two of the jboss have some configuration.
one HASingletonController(A) for my own Mbean
two SingleScheduleProvider(B)
and one HaSingletonController(C) for hasingleton-deploy
Case1
Stary "node1",
Start "node2" after "node1" is ready.
Test both node1 and node2 are working.
and all master node are on "node1"
Close "node1"
Only (A) and (B) on "node2" become Master Node
and (C) is not working (I check myself in jmx-console)
Case2
Start "node2"
Start "node1" after "node2" is ready.
test both node1 and node2 are working.
all all master node are on "node2"
close "node2"
Only (A) on "node1" become Master node
and (B) and (C) are not working
Case3
Do steps in Case2
Restart "node2"
All hasingleton on "node2" are master
(A) on "node1" still is a master
two master node of "(A)"
and I notice that they not find eachother on cluster.(no info message from console)
Terrible!!!!
------------------------
JBoss 3.2.6
Sun JDK 1.5
Hardware of "Node2" is faster then "node1"
node2: windows xp pro
node1: windows 2000 server
It is bug of misconfiguration?
PS.
"Farm" feature are not used