elasticsearch集群master节点无法重新选举

Elasticsearch | 作者 blue_sky | 发布于2020年05月18日 | 阅读数:3702

环境说明:
192.168.1.62: master data
192.168.1.66: master data
192.168.1.63: data

配置文件:
1.62:
cluster.name: es_server
node.name: node-1
network.host: 192.168.1.62
http.port: 9200
cluster.initial_master_nodes: ["node-6", "node-1"]
bootstrap.memory_lock: false
bootstrap.system_call_filter: false
node.master: true
node.data: true
discovery.seed_hosts: ["192.168.1.66", "192.168.1.62"]
discovery.zen.minimum_master_nodes: 1

1.66:
cluster.name: es_server
node.name: node-6
network.host: 192.168.1.66
http.port: 9200
cluster.initial_master_nodes: ["node-6", "node-1"]
bootstrap.memory_lock: false
bootstrap.system_call_filter: false
node.master: true
node.data: true
discovery.seed_hosts: ["192.168.1.66", "192.168.1.62"]
discovery.zen.minimum_master_nodes: 1

1.63:
cluster.name: es_server
node.name: node-2
network.host: 192.168.1.63
http.port: 9200
cluster.initial_master_nodes: ["node-6", "node-1"]
bootstrap.memory_lock: false
bootstrap.system_call_filter: false
node.master: false
node.data: true
discovery.seed_hosts: ["192.168.1.66", "192.168.1.62"]
discovery.zen.minimum_master_nodes: 1

想要配置有三个结点的集群,两个节点具有master资格,一个master node挂掉后另外一个node也可以成为master使集群继续工作。
集群启动正常可以选举出master运行正常,然后手动挂掉master node,发现另外一个具有master资格的node无法成为master,报出警告:
[node-1] master not discovered or elected yet, an election requires a node with id [fNTF4MvpQOmvkiTUeDug7A], have discovered [{node-1}{iFYOCH33Qa6sM4N1mWAifQ}{8OxfwkJ1RYaA77o1hYvhkw}{192.168.1.62}{192.168.1.62:9300}{dilm}{ml.machine_memory=8254783488, xpack.installed=true, ml.max_open_jobs=20}] which is not a quorum; discovery will continue using [192.168.1.66:9300] from hosts providers and [{node-6}{fNTF4MvpQOmvkiTUeDug7A}{eIhqTy2pQaGPaa1R2vmDTg}{192.168.1.66}{192.168.1.66:9300}{dilm}{ml.machine_memory=8254783488, ml.max_open_jobs=20, xpack.installed=true}, {node-1}{iFYOCH33Qa6sM4N1mWAifQ}{8OxfwkJ1RYaA77o1hYvhkw}{192.168.1.62}{192.168.1.62:9300}{dilm}{ml.machine_memory=8254783488, xpack.installed=true, ml.max_open_jobs=20}] from last-known cluster state; node term 1, last-accepted version 22 in term 1
已邀请:

zqc0512 - andy zhou

赞同来自: zzjxxpro

你3个节点,都搞成可以当做master吧。

trycatchfinal

赞同来自: blue_sky

7.0 以后,discovery.zen.minimum_master_nodes 这个配置就没有了。
看一下这个文章:Elasticsearch 7.0 Zen2 开启Elasticsearch分布式新纪元

blue_sky

赞同来自:

版本7.6.2

byx313 - BLOG:https://www.jianshu.com/u/43fd06f9589c

赞同来自:

"一个master node挂掉后另外一个node也可以成为master使集群继续工作"
两主的集群一台master-eligible挂掉以后集群就选不出来master了,为了防止脑裂。

zzjxxpro

赞同来自:

请问怎么解决,我有两个节点es,将其全部停掉,再分别重启之后就报master not discovered or elected yet, an election requires a node with id

要回复问题请先登录注册