Mar 14 21:27:59 lossoe1 heartbeat: [9038]: WARN: node lossoe2: is dead
Mar 14 21:27:59 lossoe1 heartbeat: [9038]: info: Dead node lossoe2 gave up resources.
Mar 14 21:27:59 lossoe1 heartbeat: [9038]: info: Link lossoe2:eth3 dead.
Mar 14 21:27:59 lossoe1 ipfail: [9066]: info: Status update: Node lossoe2 now has status dead
Mar 14 21:28:00 lossoe1 ipfail: [9066]: info: NS: We are still alive!
Mar 14 21:28:00 lossoe1 ipfail: [9066]: info: Link Status update: Link lossoe2/eth3 now has status dead
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: CRIT: Cluster node lossoe2 returning after partition.
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: info: For information on cluster partitions, See URL: http://linux-ha.org/wiki/Split_Brain
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: WARN: Deadtime value may be too small.
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: info: See FAQ for information on tuning deadtime.
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: info: URL: http://linux-ha.org/wiki/FAQ#Heavy_Load
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: info: Link lossoe2:eth3 up.
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: WARN: Late heartbeat: Node lossoe2: interval 32020 ms
Mar 14 21:28:01 lossoe1 heartbeat: [9038]: info: Status update for node lossoe2: status active
harc[73439]: 2014/03/14_21:28:01 info: Running /etc/ha.d//rc.d/status status
Mar 14 21:28:01 lossoe1 ipfail: [9066]: info: Asking other side for ping node count.
Mar 14 21:28:01 lossoe1 ipfail: [9066]: info: Checking remote count of ping nodes.
Mar 14 21:28:01 lossoe1 ipfail: [9066]: info: Link Status update: Link lossoe2/eth3 now has status up
Mar 14 21:28:01 lossoe1 ipfail: [9066]: info: Status update: Node lossoe2 now has status active
Mar 14 21:28:03 lossoe1 ipfail: [9066]: info: No giveup timer to abort.
Mar 14 21:28:03 lossoe1 heartbeat: [9038]: info: Heartbeat shutdown in progress. (9038)
Mar 14 21:28:03 lossoe1 heartbeat: [73456]: info: Giving up all HA resources.
.....
上面是我的日志输出文件，是什么问题导致切换到备用服务器上去了呢？
补充一下 我的ha.cf配置
logfile /var/log/ha_log/ha-log.log
logfacility local0
bcast eth3
keepalive 2
#warntime 10
deadtime 30
initdead 120
ucast eth3 192.168.3.1
udpport 694
auto_failback off
node lossoe1
node lossoe12
ping 10.160.0.126
respawn hacluster /usr/lib64/heartbeat/ipfail
