192.168.209.168

192.168.209.169

192.168.209.170

cp /usr/elasticsearch-6.2.3/config/elasticsearch.yml /usr/elasticsearch-6.2.3/config/elasticsearch.yml.bak

vi /usr/elasticsearch-6.2.3/config/elasticsearch.yml

cluster.name: ES_Cluster_Pcdog

node.name: 192.168.209.168

path.data: /usr/elasticsearch-6.2.3/data

path.logs: /usr/elasticsearch-6.2.3/logs

network.host: 192.168.209.168

node.master: true

node.data: true

discovery.zen.ping.unicast.hosts: ["192.168.209.168","192.168.209.169","192.168.209.170"]

discovery.zen.minimum_master_nodes: 3

discovery.zen.fd.ping_timeout: 120s

discovery.zen.fd.ping_retries: 6

discovery.zen.fd.ping_interval: 30s

cluster.routing.allocation.cluster_concurrent_rebalance: 40

cluster.routing.allocation.node_concurrent_recoveries: 40

cluster.routing.allocation.node_initial_primaries_recoveries: 40

bootstrap.memory_lock: false

bootstrap.system_call_filter: false

檢查配置，過濾掉註釋

cat elasticsearch.yml | grep -v "^#"

node 168 啟動

./bin/elasticssearch

集群狀態yellow，因為其他2臺還沒起。。。

[2018-04-20T11:36:16,000][INFO ][o.e.c.r.a.AllocationService] [192.168.209.168] Cluster health status changed from [RED] to [YELLOW] (reason: [shards started [[logstash-2018.04.18][1]] ...]).

配置文件復制到其他2個節點

scp elasticsearch.yml 192.168.209.169:/usr/elasticsearch-6.2.3/config

scp elasticsearch.yml 192.168.209.170:/usr/elasticsearch-6.2.3/config

然後在各自的node啟動elasticssearch

node 002

node 003

在master node的目錄/usr/elasticsearch-6.2.3/logs

可以看到cluster日誌

tail -f ES_Cluster_Pcdog.log

尼瑪日誌毫無更新，去其他2個節點看，2個node都有報錯，難道是我克隆出來的機器的問題？ 容我修個錯誤。

[2018-04-20T11:49:10,794][INFO ][o.e.d.z.ZenDiscovery ] [192.168.209.169] failed to send join request to master [{192.168.209.168}{-aU5102ETMW8isf85FWEHA}{V4flu0jLRfixBA0Yf9740w}{192.168.209.168}{192.168.209.168:9300}], reason [RemoteTransportException[[192.168.209.168][192.168.209.168:9300][internal:discovery/zen/join]]; nested: IllegalArgumentException[can‘t add node {192.168.209.169}{-aU5102ETMW8isf85FWEHA}{1IesX0gaSZWz2_O3eKDvfw}{192.168.209.169}{192.168.209.169:9300}, found existing node {192.168.209.168}{-aU5102ETMW8isf85FWEHA}{V4flu0jLRfixBA0Yf9740w}{192.168.209.168}{192.168.209.168:9300} with the same id but is a different node instance]; ]

原來是data裏面有原來的數據，幹掉

[pactera@ELK_002 data]$ pwd

/usr/elasticsearch-6.2.3/data

[pactera@ELK_002 data]$ rm -rf nodes/

3個節點都正常了

node 168 ，發現master 角色169

node 169 ，狀態黃變綠

node 170 ，發現master 角色169

稍後我需要研究下如何選舉master過程，這個cluster 3個master 3個data，實際生產應該是2個master然後多個data的。

ELK 實驗（六）elasticsearch集群搭建