oracle rac集群故障蓝屏重启，求解。

coologin

ORACLE RAC：由三台ＰＣ　ＳＥＲＶＥＲ组成，ORACLE 10g,WINDOWS2003,
两台　ＨＰＤＬ３８０Ｇ５,双处理器，４Ｇ内存。ＨＢＡ卡为４Ｇ。
一台　ＨＰＤＬ５７０Ｇ２,四处理器，４Ｇ内存。ＨＢＡ卡为２Ｇ。
存储：HP EVA4400

平常工作下每台内存占用２.４Ｇ，ＣＰＵ１０%。


自０７年集群７年来不间断重启，２０１１年的统计结果显示共计20次，，每台服务器都会重启，无明显规律。

心跳线是直接连接核心网上的交换机，由于集群是２００３年布置的，到现在有快十年了，网络是否是瓶颈??

心跳线是否需要提供　CACHE FUSION通信?需要高速的通信吗？


通过　ＤＵＭＰ文件发现如下：


Name
--------
=== ODM Data Collection ===

ocssd reboot the node due to 2of3 voting disk were not reachable

FileName
----------------
ocssd.log of node2

FileComment
----------------------
[ CSSD]2012-01-12 14:21:00.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 5531 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:00.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 6453 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:01.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 4531 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:01.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 5453 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:02.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 3531 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:02.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 4453 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:03.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 2531 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:03.270 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 3453 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:03.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 2515 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:03.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 3437 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:03.286 [4092] >TRACE: clssnmDiskPMT: offline disk (-2135307061 ms) (3/DGSFDB)
[ CSSD]2012-01-12 14:21:04.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 1515 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:04.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 2437 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:04.426 [224] >TRACE: clssnmSendingThread: sending status msg to all nodes
[ CSSD]2012-01-12 14:21:04.426 [224] >TRACE: clssnmSendingThread: sent 5 status msgs to all nodes
[ CSSD]2012-01-12 14:21:05.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 515 ms, disk (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:05.286 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 1437 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:05.817 [4092] >TRACE: clssnmDiskPMT: stale disk (200016 ms) (1/\\.\VOTEDSKA)
[ CSSD]2012-01-12 14:21:05.817 [4092] >WARNING: clssnmDiskPMT: voting device hang at 90% fatal, termination in 906 ms, disk (2/\\.\VOTEDSKB)
[ CSSD]2012-01-12 14:21:05.817 [4092] >ERROR: clssnmDiskPMT: Aborting, 2 of 3 voting disks unavailable
[ CSSD]2012-01-12 14:21:05.817 [4092] >ERROR: ###################################
[ CSSD]2012-01-12 14:21:05.817 [4092] >ERROR: clssscExit: CSSD aborting from thread clssnmvDiskPingMonitorThread
[ CSSD]2012-01-12 14:21:05.817 [4092] >ERROR: ###################################

coologin

组网图

一直没明白cache fusion是否通过心跳线进行通信。
节点是如何控制VOTING DISK

duolanshizhe

crs无法访问voting disk   导致crs 退出

renxiao2003

学习了。有机会得装个实验环境了。

zhlin0054

向各位学习