I'm going to setup redundant failover Redmine:

• another instance was installed on the second server without problem
• MySQL (running on the same machine with Redmine) was configured as master-master replication

Because they are in different subnet (192.168.3.x and 192.168.6.x), it seems that VIPArip is the only choice.

/etc/ha.d/ha.cf on node1

logfacility none
debug 1
debugfile /var/log/ha-debug
logfile /var/log/ha-log
autojoin none
warntime 3
deadtime 6
initdead 60
udpport 694
ucast eth1 node2.ip
keepalive 1
node node1
node node2
crm respawn

/etc/ha.d/ha.cf on node2:

logfacility none
debug 1
debugfile /var/log/ha-debug
logfile /var/log/ha-log
autojoin none
warntime 3
deadtime 6
initdead 60
udpport 694
ucast eth0 node1.ip
keepalive 1
node node1
node node2
crm respawn

crm configure show:

node $id="6c27077e-d718-4c82-b307-7dccaa027a72" node1
node $id="740d0726-e91d-40ed-9dc0-2368214a1f56" node2
primitive VIPArip ocf:heartbeat:VIPArip \
        params ip="192.168.6.8" nic="lo:0" \
        op start interval="0" timeout="20s" \
        op monitor interval="5s" timeout="20s" depth="0" \
        op stop interval="0" timeout="20s" \
        meta is-managed="true"
property $id="cib-bootstrap-options" \
        stonith-enabled="false" \
        dc-version="1.0.12-unknown" \
        cluster-infrastructure="Heartbeat" \
        last-lrm-refresh="1338870303"

crm_mon -1:

============
Last updated: Tue Jun  5 18:36:42 2012
Stack: Heartbeat
Current DC: node2 (740d0726-e91d-40ed-9dc0-2368214a1f56) - partition with quorum
Version: 1.0.12-unknown
2 Nodes configured, unknown expected votes
1 Resources configured.
============

Online: [ node1 node2 ]

 VIPArip    (ocf::heartbeat:VIPArip):   Started node1

ip addr show lo:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet 192.168.6.8/32 scope global lo
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever

I can ping 192.168.6.8 from node1 (192.168.3.x):

# ping -c 4 192.168.6.8
PING 192.168.6.8 (192.168.6.8) 56(84) bytes of data.
64 bytes from 192.168.6.8: icmp_seq=1 ttl=64 time=0.062 ms
64 bytes from 192.168.6.8: icmp_seq=2 ttl=64 time=0.046 ms
64 bytes from 192.168.6.8: icmp_seq=3 ttl=64 time=0.059 ms
64 bytes from 192.168.6.8: icmp_seq=4 ttl=64 time=0.071 ms

--- 192.168.6.8 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3000ms
rtt min/avg/max/mdev = 0.046/0.059/0.071/0.011 ms

but cannot ping virtual IP from node2 (192.168.6.x) and outside. Did I miss something?

PS: you probably want to set IP2UTIL=/sbin/ip in the /usr/lib/ocf/resource.d/heartbeat/VIPArip resource agent script if you get something like this:

Jun 5 11:08:10 node1 lrmd: [19832]: info: RA output: (VIPArip:stop:stderr) 2012/06/05_11:08:10 ERROR: Invalid OCF_RESK EY_ip [192.168.6.8]

http://www.clusterlabs.org/wiki/Debugging_Resource_Failures

Reply to @DukeLion:

Which router receives RIP updates?

When I start the VIPArip resource, ripd was run with below configuration file (on node1):

/var/run/resource-agents/VIPArip-ripd.conf:

hostname ripd
password zebra
debug rip events
debug rip packet
debug rip zebra
log file /var/log/quagga/quagga.log
router rip
!nic_tag
 no passive-interface lo:0
 network lo:0
 distribute-list private out lo:0
 distribute-list private in lo:0
!metric_tag
 redistribute connected metric 3
!ip_tag
access-list private permit 192.168.6.8/32
access-list private deny any