elasticsearch--（一）

1. 启动运行 

解压elasticsearch-1.7.2.tar.gz 后进入bin目录 

启动 ./elasticsearch -d  后台启动

访问：http://10.10.10.73:9200/

 

 vim elasticsearch.yml

(1) cluster.name  集群名称

(2) node.name  节点名称

(3) network.host  本机ip

(4) 添加脑裂配置 

discovery.zen.ping.multicast.enabled: false

discovery.zen.ping_timeout: 120s

client.transport.ping_timeout: 60s

discovery.zen.ping.unicast.hosts: ["192.168.57.4","192.168.57.5", "192.168.57.6"]

(5)启动

• 这是出于系统安全考虑设置的条件。由于ElasticSearch可以接收用户输入的脚本并且执行，为了系统安全考虑， 
  建议创建一个单独的用户用来运行ElasticSearch

• 创建elsearch用户组及elsearch用户

  groupadd elsearch
  useradd elsearch -g elsearch -p elasticsearch

• 更改elasticsearch文件夹及内部文件的所属用户及组为elsearch:elsearch

  cd /opt
  chown -R elsearch:elsearch  elasticsearch

• 切换到elsearch用户再启动

  su elsearch cd elasticsearch/bin
  ./elasticsearch

2. 安装kibana 插件

解压kibana-4.4.1-linux-x64 进入/home/insbook/kibana-4.4.1-linux-x64/config

vim kibana.xml

elasticsearch.url: "http://10.10.10.75:9200"  指向集群中的任意节点即可

(1) bin/kibana plugin --install elasticsearch/marvel/latest

 

3. elasticsearch-2.2.0 插件安装

(1) bin/plugin install license

(2) bin/plugin install marvel-agent

 

4. 2、3步骤后重启es 和kibana

 

 

2.  安装head 插件

（1）bin目录下命令安装 ./plugin -install mobz/elasticsearch-head

（2）手动安装

 

 

 

3. Bigdesk

 bin目录下命令安装 ./plugin -install lukas-vlcek/bigdesk/2.5.0

访问：http://10.10.10.73:9200/_plugin/bigdesk/#nodes

 

4. Mavel 插件

一个管理监控工具，集head和bigdesk优点为一身 收费

 bin目录下命令安装 ./plugin -install elasticsearch/marvel/latest

访问：http://10.10.10.73:9200/_plugin/bigdesk/#nodes

 

5. 索引维护

GET _search
{
  "query": {
    "match_all": {}
  }
} 

#初始化索引


PUT http://10.10.10.73:9200/y2
{
  "settings": {
    "index":{
      "number_of_shars":5,
      "mumber_of_replicas":1
    }
  }
}

GET /y2/_settings

GET /y2,y1/_settings

GET /_all/_settings



PUT /library/books/2
{
  "title":"t2",
  "name":{
    "first":"f2",
    "last":"l2"
  },
  "publist_data":"2017-15-78",
  "price":"50"
}

PUT /library/books
{
  "title":"t5",
  "name":{
    "first":"f5",
    "last":"l5"
  },
  "publist_data":"2017-15-78",
  "price":"50"
}

GET /library/books/1

GET /library/books/1?_source=title

GET /library/books/1?_source=title,price

GET /library/books/1?_source=name


POST /library/books/1/_update
{
  "doc":{
    "first":"f11"
  }
}


DELETE /library/books/1

批量获取

GET /_mget
{
  "docs" : [
      {
        "_index":"library",
        "_type":"books",
        "_id":2
      },
      {
        "_index":"library",
        "_type":"books",
        "_id":3
      }
   ]
}


GET /_mget
{
  "docs" : [
      {
        "_index":"library",
        "_type":"books",
        "_id":2,
        "_source":"title"
      },
      {
        "_index":"library",
        "_type":"books",
        "_id":3,
        "_source":"price"
      }
   ]
}


GET /_mget
{
  "docs" : [
      {
        "_index":"library",
        "_type":"books",
        "_id":2,
        "_source":"title"
      }, 
      {
        "_index":"library",
        "_type":"books",
        "_id":3,
        "_source":["price","title"]
      }
   ]
}



#获取相同index相同type下不同id的文档

GET /library/books/_mget
{
  "docs":[
      {"_id":2},
      {"_id":3}
    ]
}

#也可以简写

GET /library/books/_mget
{
  "ids":["2","5"]
}

 

批量操作

POST /library/books/_bulk
{"index":{"_id":10}}
{"title":"yuwen","price":10}
{"index":{"_id":11}}
{"title":"shuxue","price":11}
{"index":{"_id":12}}
{"title":"yingyu","price":12}
{"index":{"_id":13}}
{"title":"wuli","price":13}




GET /library/books/_mget
{
  "ids":["10","11","12","13"]
}

 

POST /library/books/_bulk
{"delete":{"_index":"library","_type":"books","_id":"10"}}
{"create":{"_index":"music","_type":"classical","_id":"14"}}
{"title":"love you"}
{"index":{"_index":"music","_type":"classical","_id":"15"}}
{"title":"love me"}
{"update":{"_index":"library","_type":"books","_id":"11"}}
{"doc":{"price":"100"}}

 

映射的属性和方法

 

# 建立映射
POST /library
{
    "settings": {
        "number_of_shards": 5,
        "number_of_replicas": 1
    },
    "mappings": {
        "books": {
            "properties": {
                "title": { "type": "string"},
                "name": { "type": "string", "index": "not_analyzed"},
                "publish_date": {"type": "date", "index": "not_analyzed"},
                "price": {"type": "double"},
                "number": {"type": "integer"}
                }
            }
    }
}

# 动态映射
PUT /library
{
       "mappings": {
        "books": {
            "dynamic":  "strict",
            "properties": {
                "title": { "type": "string" },
                "name": { "type": "string", "index": "not_analyzed"},
                "publish_date": {"type": "date", "index": "not_analyzed"},
                "price": {"type": "double"},
                "number": {
                  "type":   "object",
                  "dynamic":  true
                  }
              }
          }
    }
   
}

GET /library/

# 获取某个索引的映射信息
GET /library/_mapping

GET /bank/_mapping/

# 获取某个索引下某个type的映射信息
GET /library/_mapping/books

# 获取这个集群内所有的映射信息
GET /_all/_mapping

# 获取这个集群内某两个或多个type的映射信息
GET /_all/_mapping/books,bank_account


# 更新修改Mapping映射
# 很遗憾，mapping一旦建立，就不能修改现有的字段映射
# 如果要推倒现有的映射，你得重新建立一个索引，然后重新定义映射
# 然后把之前索引里的数据导入到新建立的索引里
# ---------具体的方法----------
# 1. 给现有的索引定义一个别名，并且把现有的索引指向这个别名，运行步骤2
# 2. 运行 ： PUT /现有索引/_alias/别名A
# 3. 新创建一个索引，定义好最新的映射
# 4. 将别名指向新的索引，并且取消之前索引的指向，运行步骤5
# 5. 运行： POST /_aliases
#  {
#    "actions": [
#        { "remove": { "index": "现有索引名", "alias": "别名A" }},
#        { "add":    { "index": "新建索引名", "alias": "别名A" }}
#    ]
#  }
#
# 注：通过这几个步骤就实现了索引的平滑过度，并且是零停机的
#


#删除映射
DELETE /library/books

DELETE /library/books/_mapping

DELETE /library/_mapping/books,bank