下面是节点galera 节点197.3.155.152异常关闭之后，197.3.155.151节点的日志：

2016-07-07 11:05:49 1400117****4240 [Note] WSREP: (fb55d9c9, 'tcp://0.0.0.0:4567') turning message relay requesting on, nonlive peers: tcp://197.3.155.152:4567 

发现nonlive的节点197.3.155.152

2016-07-07 11:05:51 1400117****4240 [Note] WSREP: (fb55d9c9, 'tcp://0.0.0.0:4567') reconnecting to 51f4c07a (tcp://197.3.155.152:4567), attempt 0

重新尝试发送一个keepalived的链接。

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: evs::proto(fb55d9c9, GATHER, view_id(REG,128a493c,122)) suspecting node: 51f4c07a

经过5秒还没有收到响应，怀疑节点已经宕掉，51f4c07a即197.3.155.152。

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: evs::proto(fb55d9c9, GATHER, view_id(REG,128a493c,122)) suspected node without join message, declaring inactive

宣判节点宕掉。declaring inactive

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: declaring 128a493c at tcp://197.3.155.150:4567 stable

获取存活的节点信息。

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: Node 128a493c state prim

获取存活的节点的状态。

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: view(view_id(PRIM,128a493c,123) memb {

        128a493c,0

        fb55d9c9,0

} joined {

} left {

} partitioned {

        51f4c07a,0

})

对宕掉的节点进行移除，即partitioned

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: save pc into disk

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: forgetting 51f4c07a (tcp://197.3.155.152:4567)

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: New COMPONENT: primary = yes, bootstrap = no, my_idx = 1, memb_num = 2

打印新的primary COMPONENT 信息。

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: STATE EXCHANGE: Waiting for state UUID.

2016-07-07 11:05:55 1400117****4240 [Note] WSREP: (fb55d9c9, 'tcp://0.0.0.0:4567') turning message relay requesting off

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: STATE EXCHANGE: sent state msg: b8a08882-43ef-11e6-884e-2beaea270cdb

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: STATE EXCHANGE: got state msg: b8a08882-43ef-11e6-884e-2beaea270cdb from 0 (node3)

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: STATE EXCHANGE: got state msg: b8a08882-43ef-11e6-884e-2beaea270cdb from 1 (node2)

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: Quorum results:

        version    = 4,

        component  = PRIMARY,

        conf_id    = 33,

        members    = 2/2 (joined/total),

        act_id     = 128****12,

        last_appl. = 0,

        protocols  = 0/7/3 (gcs/repl/appl),

        group UUID = afbfeed1-38e8-11e6-84b1-d20f5b484535

新的集群状态信息在新的primary component 内的节点进行交换与确认。

2016-07-07 11:05:55 1400117****1536 [Note] WSREP: Flow-control interval: [23, 23]

2016-07-07 11:05:55 1400119****1184 [Note] WSREP: New cluster view: global state: afbfeed1-38e8-11e6-84b1-d20f5b484535:128****12, view# 34: Primary, number of nodes: 2, my index: 1, protocol version 3

2016-07-07 11:05:55 1400119****1184 [Note] WSREP: wsrep_notify_cmd is not defined, skipping notification.

2016-07-07 11:05:55 1400119****1184 [Note] WSREP: REPL Protocols: 7 (3, 2)

2016-07-07 11:05:55 1400118****6944 [Note] WSREP: Service thread queue flushed.

2016-07-07 11:05:55 1400119****1184 [Note] WSREP: Assign initial position for certification: 128****12, protocol version: 3

2016-07-07 11:05:55 1400118****6944 [Note] WSREP: Service thread queue flushed.

2016-07-07 11:05:58 1400117****4240 [Note] WSREP:  cleaning up 51f4c07a (tcp://197.3.155.152:4567)

完全剔除197.3.155.152。

上面的日志中，涉及到进行节点宕机检查的timeout 时间，详细信息请参考官方文档，作者做了截图，请查看！

几个相关的参数有相对大小的关系，特殊设置时请注意：

顺便提一下，在galera 集群中，执行ddl会造成整个集群hang ,如果ddl只要执行的时间较长，在交易频率的系统上，则就造成了悲剧。下面的方式可以避免这种情况。

To run an ALTER statement in this manner, on each node run the following queries:

在galera集群的每一个节点上执行下面的步骤：

1. Change the Schema Upgrade method to Rolling Schema Upgrade.

SET wsrep_OSU_method='RSU';

设置为滚动schema变更模式。

2. Run the ALTER statement.

执行alter 语句。

3. Reset the Schema Upgrade method back to Total Order Isolation.

SET wsrep_OSU_method='TOI';

将schema的变更模式重新设置为TOI。