elasticsearch-6.1.2安装配置维护

╰+哭是因爲堅強的太久メ 2022-05-11 12:52 274阅读 0赞

参考手册：https://cloud.tencent.com/developer/article/1011050  
es版本：  
1. 3台6.1.2的版本的作ES集群用于日志和报表,安装插件如下:  
a. elasticsearch-analysis-ik 6.1.2  
b. Elasticsearch-head 6.1.2 可视化管理平台

系统环境：centos7.5  
es版本：elasticsearch-6.1.2  
jdk：1.8  
主机ip  
192.168.2.126  
192.168.2.127  
192.168.2.128

1、软件下载  
wget https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-6.1.2.tar.gz

2、jdk安装  
wget ftp://10.249.100.249/pub/jdk/jdk8/\*

3、安装es

tar -xzvf elasticsearch-6.1.2.tar.gz -C /usr/local/  
cd /usr/local/  
mv elasticsearch-6.1.2/ elasticsearch

4、建立data目录和log目录  
mkdir -p /data/elasticsearch/data  
mkdir -p /data/elasticsearch/logs

5、建立用户(es不能用root运行)  
useradd es  
chown -R es:es /usr/local/elasticsearch  
chown -R es:es /data

6、系统参数优化  
vim /usr/local/elasticsearch/config/jvm.options  
\-Xms4g  
\-Xmx4g

vi /etc/sysctl.conf  
   vm.max\_map\_count=262144  
   fs.file-max=655360

vi /etc/security/limits.conf  
    es         soft   memlock      unlimited  
    es        hard   memlock      unlimited

7、编辑配置文件  
\[root@ip-192-168-202-126 ~\]\# grep '^\[a-z\]' /usr/local/elasticsearch/config/elasticsearch.yml  
cluster.name: es6.1.2  
node.name: node-126  
node.master: true  
path.data: /data/elasticsearch/data  
path.logs: /data/elasticsearch/logs  
bootstrap.memory\_lock: true  
network.host: 0.0.0.0  
http.port: 9200  
transport.tcp.port: 9300  
discovery.zen.ping.unicast.hosts: \["192.168.2.126", "192.168.2.127", "192.168.2.128"\]  
gateway.recover\_after\_nodes: 3

\[root@ip-192-168-202-127 bin\]\# grep '^\[a-z\]' /usr/local/elasticsearch/config/elasticsearch.yml  
cluster.name: es6.1.2  
node.name: node-127  
node.master: true  
path.data: /data/elasticsearch/data  
path.logs: /data/elasticsearch/logs  
bootstrap.memory\_lock: true  
network.host: 0.0.0.0  
http.port: 9200  
transport.tcp.port: 9300  
discovery.zen.ping.unicast.hosts: \["192.168.2.126", "192.168.2.127", "192.168.2.128"\]  
gateway.recover\_after\_nodes: 3

\[root@ip-192-168-202-128 ~\]\# grep '^\[a-z\]' /usr/local/elasticsearch/config/elasticsearch.yml  
cluster.name: es6.1.2  
node.name: node-128  
node.master: true  
path.data: /data/elasticsearch/data  
path.logs: /data/elasticsearch/logs  
bootstrap.memory\_lock: true  
network.host: 0.0.0.0  
http.port: 9200  
transport.tcp.port: 9300  
discovery.zen.ping.unicast.hosts: \["192.168.2.126", "192.168.2.127", "192.168.2.128"\]  
discovery.zen.minimum\_master\_nodes: 2  
gateway.recover\_after\_nodes: 3

8、启动  
su - es  
/usr/local/elasticsearch/bin/elasticsearch -d

查找ES进程  
ps -ef | grep elastic  
杀掉ES进程  
kill -9 2382（进程号）  
重启ES  
sh elasticsearch -d

9、检查  
\[es@ip-192-168-202-126 ~\]$ curl localhost:9200  
\{  
  "name" : "node-126",  
  "cluster\_name" : "es6.1.2",  
  "cluster\_uuid" : "-k1nGSoAQri3ToEqSoQHYg",  
  "version" : \{  
    "number" : "6.1.2",  
    "build\_hash" : "5b1fea5",  
    "build\_date" : "2018-01-10T02:35:59.208Z",  
    "build\_snapshot" : false,  
    "lucene\_version" : "7.1.0",  
    "minimum\_wire\_compatibility\_version" : "5.6.0",  
    "minimum\_index\_compatibility\_version" : "5.0.0"  
  \},  
  "tagline" : "You Know, for Search"  
\}  
\[es@ip-192-168-202-126 ~\]$ jps  
14817 Elasticsearch  
15928 Jps  
\[es@ip-192-168-202-127 ~\]$ curl localhost:9200  
\{  
  "name" : "node-127",  
  "cluster\_name" : "es6.1.2",  
  "cluster\_uuid" : "-k1nGSoAQri3ToEqSoQHYg",  
  "version" : \{  
    "number" : "6.1.2",  
    "build\_hash" : "5b1fea5",  
    "build\_date" : "2018-01-10T02:35:59.208Z",  
    "build\_snapshot" : false,  
    "lucene\_version" : "7.1.0",  
    "minimum\_wire\_compatibility\_version" : "5.6.0",  
    "minimum\_index\_compatibility\_version" : "5.0.0"  
  \},  
  "tagline" : "You Know, for Search"  
\}  
\[es@ip-192-168-202-127 ~\]$ jps  
15026 Elasticsearch  
16091 Jps

es@ip-192-168-202-128 ~\]$ curl localhost:9200  
\{  
  "name" : "node-128",  
  "cluster\_name" : "es6.1.2",  
  "cluster\_uuid" : "-k1nGSoAQri3ToEqSoQHYg",  
  "version" : \{  
    "number" : "6.1.2",  
    "build\_hash" : "5b1fea5",  
    "build\_date" : "2018-01-10T02:35:59.208Z",  
    "build\_snapshot" : false,  
    "lucene\_version" : "7.1.0",  
    "minimum\_wire\_compatibility\_version" : "5.6.0",  
    "minimum\_index\_compatibility\_version" : "5.0.0"  
  \},  
  "tagline" : "You Know, for Search"  
\}  
\[es@ip-192-168-202-128 ~\]$ jps

16033 Jps  
14968 Elasticsearch

或者登陆检查  
http://192.168.2.126:9200/  
http://192.168.2.127:9200/  
http://192.168.2.128:9200/

Elasticsearch的健康检查

https://www.elastic.co/guide/en/elasticsearch/guide/current/\_cluster\_health.html

Elasticsearch提供了自身用于健康检查的API，会返回一组JSON的数据：  
curl -XGET http://192.168.2.126:9200/\_cluster/health   
curl -XGET http://192.168.2.126:9200/\_cluster/health?pretty=true

同时，官方还提供了多样监控信息，可以通过cat API的方式来返回需要的数据：  
https://www.elastic.co/guide/en/elasticsearch/guide/current/\_cat\_api.html   
curl -XGET http://192.168.2.128:9200/\_cat/health

对集群中的任意一个节点去查询，此节点都会从搜集到整个集群的信息进行返回：  
 curl -XGET http://192.168.2.128:9200/\_cat/nodes?v  
   
Elasticsearch有green/yellow/red  三种不同的状态级别，  
green 表示所有节点分片正常，yellow 表示主节点分片正常，副本节点分片存在故障或丢失，red 表示主节点分片丢失，在搜索时只能返回部分结果。  
   
\[es@ip-192-168-202-127 ~\]$ curl -XGET http://192.168.2.126:9200/\_cluster/health?pretty=true  
\{  
  "cluster\_name" : "es6.1.2",  
  "status" : "green",  
  "timed\_out" : false,  
  "number\_of\_nodes" : 3,  
  "number\_of\_data\_nodes" : 3,  
  "active\_primary\_shards" : 0,  
  "active\_shards" : 0,  
  "relocating\_shards" : 0,  
  "initializing\_shards" : 0,  
  "unassigned\_shards" : 0,  
  "delayed\_unassigned\_shards" : 0,  
  "number\_of\_pending\_tasks" : 0,  
  "number\_of\_in\_flight\_fetch" : 0,  
  "task\_max\_waiting\_in\_queue\_millis" : 0,  
  "active\_shards\_percent\_as\_number" : 100.0  
\}  
\[es@ip-192-168-202-127 ~\]$ curl -XGET http://192.168.2.128:9200/\_cat/health  
1534565131 12:05:31 es6.1.2 green 3 3 0 0 0 0 0 0 - 100.0%  
\[es@ip-192-168-202-127 ~\]$ curl -XGET http://192.168.2.128:9200/\_cat/nodes?v  
ip              heap.percent ram.percent cpu load\_1m load\_5m load\_15m node.role master name  
192.168.2.126            7          63   0    0.02    0.16     0.15 mdi       -      node-126  
192.168.2.127            7          62   0    0.00    0.07     0.08 mdi       \*      node-127  
192.168.2.128            4          63   1    0.00    0.06     0.08 mdi       -      node-128  
\[es@ip-192-168-202-127 ~\]$ curl -XGET http://192.168.2.127:9200/\_cat/nodes?v  
ip              heap.percent ram.percent cpu load\_1m load\_5m load\_15m node.role master name  
192.168.2.126            7          63   0    0.01    0.15     0.15 mdi       -      node-126  
192.168.2.127            4          62   0    0.00    0.06     0.08 mdi       \*      node-127  
192.168.2.128            4          63   0    0.00    0.06     0.08 mdi       -      node-128

==============================================  
\#集群的名称  
cluster.name: es6.2  
\#节点名称,其余两个节点分别为node-2 和node-3  
node.name: node-1  
\#指定该节点是否有资格被选举成为master节点，默认是true，es是默认集群中的第一台机器为master，如果这台机挂了就会重新选举master  
node.master: true  
\#允许该节点存储数据(默认开启)  
node.data: true  
\#索引数据的存储路径  
path.data: /usr/local/elk/elasticsearch/data  
\#日志文件的存储路径  
path.logs: /usr/local/elk/elasticsearch/logs  
\#设置为true来锁住内存。因为内存交换到磁盘对服务器性能来说是致命的，当jvm开始swapping时es的效率会降低，所以要保证它不swap  
bootstrap.memory\_lock: true  
\#绑定的ip地址  
network.host: 0.0.0.0  
\#设置对外服务的http端口，默认为9200  
http.port: 9200  
\# 设置节点间交互的tcp端口,默认是9300   
transport.tcp.port: 9300  
\#Elasticsearch将绑定到可用的环回地址，并将扫描端口9300到9305以尝试连接到运行在同一台服务器上的其他节点。  
\#这提供了自动集群体验，而无需进行任何配置。数组设置或逗号分隔的设置。每个值的形式应该是host:port或host  
\#（如果没有设置，port默认设置会transport.profiles.default.port 回落到transport.tcp.port）。  
\#请注意，IPv6主机必须放在括号内。默认为127.0.0.1, \[::1\]  
discovery.zen.ping.unicast.hosts: \["192.168.8.101:9300", "192.168.8.103:9300", "192.168.8.104:9300"\]  
\#如果没有这种设置,遭受网络故障的集群就有可能将集群分成两个独立的集群 - 分裂的大脑 - 这将导致数据丢失  
discovery.zen.minimum\_master\_nodes: 2