关于再议RAC Brain Split脑裂的问题

Maclean Liu(刘相兵

网友提问, Question:

你好！ 我看了你的这篇“再议脑裂”，文中说了判断有效集群的方法是：
   

• 拥有最多节点数目的子集群(Sub-cluster with largest number of Nodes)
• 若子集群内数目相等则为拥有最低节点号的子集群(Sub-cluster with lowest node        number)，举例来说在一个2节点的RAC环境中总是1号节点会获胜；

    现在，我有一个问题：假如子集群中的节点数相同，而且最小节点号都是1，这时候需要怎么处理？是继续比较子集群中其他的节点号，还是有其他方法    吗？

Maclean Liu(刘相兵

http://www.oracledatabase12g.com ... lit-resolution.html

The RM (Reconfig Manager) sends a sync message to all participating nodes. Participating nodes respond with a sync acknowledgement. After this the vote phase begins and the master sends a vote message to all participating nodes. Participating nodes repond with a vote info message containing their node identifier and GM peer to peer listening endpoint. In the split-check phase, the RM uses the voting disk to verify there is no split-brain. It finds nodes heartbeating to disk that are not connected via the network. If it finds these, it will determine which nodes are talking to which and the largest subcluster survives.


For example, if we have a 5 node cluster and all of the nodes are heartbeating to the voting disk but only a group of 3 can communicate via the network and a group of 2 can communication via the network, this means we have 2 subclusters. The largest subcluster (3) would survive while the other subcluster (2) would not. After this the evict phase would evict nodes previously in the cluster but not considered members in this incarnation. In this case we would send a message to evicted nodes (if possible) and write eviction notice to a ‘kill’ block in the voting file. We would wait for the node to indicate it got the eviction notice (wait for seconds).


The wait is terminated by a message or status on the voting file indicating that the node got the eviction notice. In the update phase the master sends an update message containing the definitive cluster membership and node information for all particpating nodes. The participating nodes send update acknowledgements. All members queue the reconfiguration event ot their GM.

panfengfeng

还有一个就是RAC中判断有效集群是通过一个master去判断的还是说故障节点自身判断？
PS：最近我在研究oracle的分布式文件系统——ocfs2，有一些机制不怎么明白，在ocfs2中，判断脑裂是由故障节点自身判断，判断规则大致是：       
1、它能看到奇数个有心跳的节点,并且与大于等于一半的节点有网络连接；
2、        它能看到偶数个有心跳的节点,至少与一半节点有网络连接,并且与node number号最小的节点有连接,即判定能连接  上节点号最小的节点的集群为有效集群；
不知道您对ocfs2 这块有没有什么看法？

Maclean Liu(刘相兵

Answer:

“假如子集群中的节点数相同，而且最小节点号都是1”

RAC 中不存在 2个 节点的节点号一样的情况， 即不存在最小节点号都是1的可能性。


Some Metadata LOG


Node2 :

[root@vrh2 cssd]# cat ocssd.l01 |grep clssgmCMReconfig
2011-12-04 22:10:07.517: [    CSSD][1127934272]clssgmCMReconfig: reconfiguration successful, incarnation 216157713 with 1 nodes, local node number 2, master node number 2
2011-12-05 22:00:47.269: [    CSSD][1127532864]clssgmCMReconfig: reconfiguration successful, incarnation 216157715 with 1 nodes, local node number 2, master node number 2
2011-12-05 22:03:07.763: [    CSSD][1127532864]clssgmCMReconfig: reconfiguration successful, incarnation 216157716 with 2 nodes, local node number 2, master node number 2
2011-12-06 04:48:24.839: [    CSSD][1127721280]clssgmCMReconfig: noticed master(1) not supporting REPLAY, cleaning up grocks autonomously
2011-12-06 04:48:25.253: [    CSSD][1127721280]clssgmCMReconfig: reconfiguration successful, incarnation 216157718 with 2 nodes, local node number 2, master node number 1
2011-12-13 09:22:32.767: [    CSSD][1129257280]clssgmCMReconfig: noticed master(1) not supporting REPLAY, cleaning up grocks autonomously
2011-12-13 09:22:32.971: [    CSSD][1129257280]clssgmCMReconfig: reconfiguration successful, incarnation 216157721 with 2 nodes, local node number 2, master node number 1



Node 1:

[root@vrh1 cssd]# cat ocssd.l01 |grep clssgmCMReconfig
2011-12-13 09:22:33.200: [    CSSD][1130473792]clssgmCMReconfig: reconfiguration successful, incarnation 216157721 with 2 nodes, local node number 1, master node number 1
2011-12-18 06:34:31.890: [    CSSD][1130293568]clssgmCMReconfig: noticed master(2) not supporting REPLAY, cleaning up grocks autonomously
2011-12-18 06:34:31.932: [    CSSD][1130293568]clssgmCMReconfig: reconfiguration successful, incarnation 216157724 with 2 nodes, local node number 1, master node number 2
2011-12-19 21:40:22.468: [    CSSD][1128020288]clssgmCMReconfig: reconfiguration successful, incarnation 216157726 with 1 nodes, local node number 1, master node number 1



clssgmCMReconfig   ==> Group Management Cluster Master Reconfig  

这里可以看到 一号节点的  node number  一般是1
                        二号节点的  node number  一般是2


更多信息请参考 关于RAC 确定 master节点疑问 这个帖子：http://t.askmaclean.com/thread-180-1-1.html