elasticsearch自动按天创建索引脚本

elasticsearch保存在一个索引中数据量太大无法查询，现在需要将索引按照天来建，查询的时候关联查询即可

有时候es集群创建了很多索引，删不掉，如果是测试环境或者初始化es集群(清空所有数据)，可以直接关掉elastic进程，然后删除nodes下面的所有数据，再次启动集群即可，记录一下避免忘记

导出mapping信息放到/root/index_mapping目录下

1.导出的语句
yum install epel-release -y
yum install nodejs -y
yum install nodejs npm -y
npm install elasticdump -y

/root/node_modules/elasticdump/bin/elasticdump --ignore-errors=true --scrollTime=120m --bulk=true --input=http://10.30.138.62:9200/.kibana --output=mapping.json --type=mapping

2.创建索引的脚本如下

#!/bin/bash

# .创建今天和明天的索引
# .删除3天以前的索引

today_date=`date '+%Y%m%d'`
tomorrow_date=`date -d tomorrow +%Y%m%d`
# 需要导入索引的es集群服务器ip
es_ip=10.10.33.84

# 找到具体的index目录和index，对这个index进行处理
# start log
echo "${today_date} create index start">> /data/scripts/create_index.log
for FULLPATH in `ls /root/index_mapping/*.json`;
do
    # 文件名
    FILE=${FULLPATH##*/}
    # 去掉文件名后缀
    FILE_NAME=${FILE%%.*}
    # 找到类似push:user:req的索引名称
    FILE_INDEX=`echo $FILE_NAME|sed 's/_/:/g'`
    FILE_INDEX_NAME=${FILE_INDEX%:*}
    echo "create index ${FILE_INDEX_NAME} start" >> /data/scripts/create_index.log
    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:${today_date} --type=mapping
    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:${tomorrow_date} --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170905 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170904 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170903 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170906 --type=mapping
done
# end log
echo "${today_date} create index end">> /data/scripts/create_index.log

three_daysago=`date -d '3 days ago' +%Y%m%d`
echo $three_daysago
echo "delete index ${three_daysago} start" >> /data/scripts/create_index.log
curl -XDELETE "http://10.10.33.84:9200/*${three_daysago}*"
echo "delete index ${three_daysago} end" >> /data/scripts/create_index.log

某些索引特殊处理，改进后的脚本

#!/bin/bash

# .创建今天和明天的索引
# .删除3天以前的索引

today_date=`date '+%Y%m%d'`
tomorrow_date=`date -d tomorrow +%Y%m%d`
today_month=`date '+%Y%m'`
# 需要导入索引的es集群服务器ip
es_ip=10.10.33.84

# 找到具体的index目录和index，对这个index进行处理
# start log
echo "${today_date} create index start">> /data/scripts/create_index.log
for FULLPATH in `ls /root/index_mapping/*.json`;
do
    # 文件名
    FILE=${FULLPATH##*/}
    # 去掉文件名后缀
    FILE_NAME=${FILE%%.*}
    # 找到类似push:user:req的索引名称
    FILE_INDEX=`echo $FILE_NAME|sed 's/_/:/g'`
    FILE_INDEX_NAME=${FILE_INDEX%:*}
    echo "create index ${FILE_INDEX_NAME} start" >> /data/scripts/create_index.log
    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:${today_date} --type=mapping
    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:${tomorrow_date} --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170905 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170904 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170903 --type=mapping
#    /root/node_modules/elasticdump/bin/elasticdump --input=${FULLPATH} --output=http://${es_ip}:9200/${FILE_INDEX_NAME}:20170906 --type=mapping
done
# end log
echo "${today_date} create index end">> /data/scripts/create_index.log

three_daysago=`date -d '3 days ago' +%Y%m%d`
echo $three_daysago
echo "delete index ${three_daysago} start" >> /data/scripts/create_index.log
# 索引保留3天
curl -XDELETE "http://${es_ip}:9200/voice*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/script*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/push*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/advert*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/user*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/speech*${three_daysago}*"
curl -XDELETE "http://${es_ip}:9200/user*${three_daysago}*"
echo "delete index ${three_daysago} end" >> /data/scripts/create_index.log
# bin开头的索引保留7天
seven_daysago=`date -d '7 days ago' +%Y%m%d`
echo $seven_daysago
echo "delete index ${seven_daysago} start" >> /data/scripts/create_index.log
curl -XDELETE "http://${es_ip}:9200/bin*${seven_daysago}*"
echo "delete index ${seven_daysago} end" >> /data/scripts/create_index.log

curl -XPUT "http://${es_ip}:9200/*${today_month}*/_settings" -d '{"number_of_replicas": 0}'
echo "setting replication 0 ${today_date}" >> /data/scripts/create_index.log

查看索引结构的命令：

[root@u04es01 ~]# curl 10.19.142.99:/bin:user:task:/_mapping?pretty
{
  "bin:user:task:20171230" : {
    "mappings" : {
      "_default_" : {
        "properties" : {
          "appId" : {
            "type" : "keyword",
            "store" : true
          },
          "taskFlag" : {
            "type" : "integer",
            "store" : true
          },
          "taskId" : {
            "type" : "keyword",
            "store" : true
          },
          "time" : {
            "type" : "date",
            "store" : true,
            "format" : "yyyy-MM-dd HH:mm:ss.SSS.Z"
          },
          "uuid" : {
            "type" : "keyword",
            "store" : true
          }
        }
      }
    }
  }
}