SQL Cluster nodes both show passive

dagudboi

Have any of you folks experienced something similar to this?  EMC SAN SP was rebooted and then both cluster nodes appeared as passive/reserve therefore no active node was online.  It has created a big outage until the a rolling reboot was performed.  Now no one knows how a SP reboot could put cluster nodes out.

刚才用公司电脑无法打中文，情况是这样的，两个刀片服务器上运行两个节点的SQL数据库集群，一天前SAN上的一个SP被重新设置了，然后两个节点都处于等待状态，也就是说没有一个服务器在线提供服务了，感觉是两个刀片都无法联系到存储上的quorum了。有没有哪位朋友遇到过相同的问题分享一下，看看到底是那里出问题了.

[ 本帖最后由 dagudboi 于 2011-3-16 22:39 编辑 ]

YugaYuga

1. power off (or shut down cluster service on one node) one node
2. on the remaining node, from cmd, run:
net start clussvc /fq
(if does not work, try: net start clussvc /fixquorum
The above hopefully will reset the quorum.
3. power on the other node

Try the above first see what happens.

yesngo

LZ的服务器都已经恢复工作了，你还要让他强制启动cluster服务。估计他老板不会答应。

dagudboi

原帖由 YugaYuga 于 2011-3-16 23:21 发表
1. power off (or shut down cluster service on one node) one node
2. on the remaining node, from cmd, run:
net start clussvc /fq
(if does not work, try: net start clussvc /fixquorum
The above hopefully ...

多谢出谋划策，不过就像楼上的朋友说的一样，问题已经解决了，我不是负责排除问题的工程师但要参与整个集群的重新设计。有趣的是出现问题的时候，两个刀片的集群服务都在运行，只是都处于待机状态。集群资源里也是空空如也。

yesngo

问题是存储引起的,那多半就是存储的问题或者是存储连接的问题.着坛子里有几个存储的大牛.要是他们能指点一下,估计会对你有帮助.