server fail safe

[server/]

What parameter to use in heartbeat configuration file so as to stablize the virtual ip address permanently on second node if first node goes down in four node architecture as ip floating between second and third when making second primary and third as backup for second.

[techstress]

whats the location of the heartbeat file?

how did you setup this system?

is it a regular cluster?

[server/]

hi,
previously i was trying to establish four node architecture using heartbeat v1 and i got the problem but now i shifted to heartbeat 2 and trying same architecture

4 node like
Node A, B, C, D.
and
one resource Ipaddr

but here i am facing problem with ip failover as

1- Node A goes down then with default setting and without constraint which node will get the ip resource ????.what will be the default behaviour of cluster in such schenario.

2- if node B gets resource in above case then when i make node A up and node B down then resource is failing back on node A which is not expected either it should move to node C or D which is desired


cib.xml

no-quorum-policy --- ignore
resource-stickiness ---- INFINITY
resource-failure-stickiness---- 0

node description
then resource description having monitor operation on

no constraints


ha.cf

crm --- yes
autojoin ---- yes
node A B C D
auto failback off


any one having information related it please help in need.......
thanks in advance