es版本 5.2.2
[2018-07-25T21:26:03,713][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][92] received shard failed for shard id [[flume-hour20-2018-07-25][92]], allocation id [YiY7_m-JQVypIXjoiPknOg], primary term [2], message [mark copy as stale]
[2018-07-25T21:26:03,714][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][65] received shard failed for shard id [[flume-hour20-2018-07-25][65]], allocation id [L_Q8UF-TQ5eM7PWMrezVJQ], primary term [2], message [mark copy as stale]
[2018-07-25T21:26:03,714][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][68] received shard failed for shard id [[flume-hour20-2018-07-25][68]], allocation id [TILQZxatSciBFuXtFMlT5w], primary term [1], message [mark copy as stale]
[2018-07-25T21:26:03,715][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][111] received shard failed for shard id [[flume-hour20-2018-07-25][111]], allocation id [t1v3JsoaRYO-R1PJNAfHNA], primary term [1], message [mark copy as stale]
[2018-07-25T21:26:03,715][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][131] received shard failed for shard id [[flume-hour20-2018-07-25][131]], allocation id [RaWA0kjjSe-feEoP6PY2mw], primary term [1], message [mark copy as stale]
[2018-07-25T21:26:03,716][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][19] received shard failed for shard id [[flume-hour20-2018-07-25][19]], allocation id [iu4707o9SkmxLVlYyD7UBA], primary term [1], message [mark copy as stale]
[2018-07-25T21:26:03,716][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][131] received shard failed for shard id [[flume-hour20-2018-07-25][131]], allocation id [RaWA0kjjSe-feEoP6PY2mw], primary term [1], message [mark copy as stale]
[2018-07-25T21:26:03,717][WARN ][o.e.c.a.s.ShardStateAction] [es1.1.master] [flume-hour20-2018-07-25][52] received shard failed for shard id [[flume-hour20-2018-07-25][52]], allocation id [HUCdOBhjTlOD73kpBDu6yQ], primary term [1], message [mark copy as stale]
EDIT:
今天好好看了下日志。 是因为挂掉的那个datanode节点进程没挂掉。 cpu load较高，但是偶尔还能ping的通，导致master没把节点t掉，所以数据写不进。 跟异常日志没有关系。 变成red，是忘记了一个老index，没有rep。出问题的时候太慌，只顾着看满屏刷不停的这个日志。
3 个回复
rochy - rochy_he
赞同来自:
你的主节点配置了几个？ 最小要求几个？
如果你的 minimum_master_nodes 设置的是2，目前主节点只有一个的话，那么整个集群是不可用的
zqc0512 - andy zhou
赞同来自:
应该可用的。
laoyang360 - 《一本书讲透Elasticsearch》作者，Elastic认证工程师 [死磕Elasitcsearch]知识星球地址：http://t.cn/RmwM3N9；微信公众号：铭毅天下; 博客：https://elastic.blog.csdn.net
赞同来自:
最小节点数是防止脑裂设置的。3节点集群：建议都设置为即是：master又是data,最小主节点数设置为2,这样一个宕机，另外两个也满足集群条件。集群可以正常运行的。