Easticsearch 数据迁移至in

#!/usr/bin/env python
#coding=utf-8

import sys
import json
import datetime
import elasticsearch
from influxdb import InfluxDBClient

#连接Easticsearch
class ES(object):
    @classmethod
    def connect_host(cls):
        url = "http://192.168.121.33:9202/"
        es = elasticsearch.Elasticsearch(url,timeout=120)
        return es
es = ES.connect_host()

#连接influxdb
client = InfluxDBClient(host="192.168.121.33", port="8086", username='admin', password='admin', database='esl')
client.create_database('esl')

#DSL查询语法
data = {
    "query": { "match_all" : {}},
    "size": 100
}

# 设置要过滤返回的字段值，要什么字段。
    'hits.hits._source.resource_id',
    'hits.hits._source.timestamp',
    'hits.hits._source.counter_volume',
    'hits.hits._source.@timestamp',
]

# 指定search_type="scan"模式，并返回_scroll_id给es.scroll获取数据使用
res = es.search(
    index='pipefilter_meters*',
    doc_type ='canaledge.flow.bytes',
    body=data,
    search_type="scan",
    scroll="10m"
)
scroll_id = res['_scroll_id']

response= es.scroll(scroll_id=scroll_id, scroll= "10m",filter_path=return_fields,)
scroll_id = response['_scroll_id']   #获取第二次scroll_id
hits = response['hits']['hits'] 
in_data = []

while len(hits) > 0:
    for i in hits:
        res_id = i['_source']['resource_id']
        r_id, r_type = res_id.split(':')
        datas = {
            "measurement": "es_net",
            "tags": {
                 "resource_id": r_id,
                 "type": r_type
             },
            "time": i['_source']['timestamp'],
            "fields": {
                "counter_volume": i['_source']['counter_volume']
            }
        }
        in_data.append(datas)
    #循环写入influxdb
    client.write_points(in_data)
    in_data = []   #每次循环完重新定义列表为空

    data = {
        "query": { "match_all" : {}},
        "size": 100
    }
    ## 设置要过滤返回的字段值，要什么字段。
        '_scroll_id',
        'hits.hits._source.resource_id',
        'hits.hits._source.timestamp',
        'hits.hits._source.counter_volume',
        'hits.hits._source.@timestamp',
    ]

    ## 指定search_type="scan"模式，并返回_scroll_id给es.scroll获取数据使用
    response= es.scroll(scroll_id=scroll_id, scroll= "10m",filter_path=return_fields,)
    #调试
    #if not response.get('hits'):
    #    print response
    #    sys.exit(1)
    #else:
    
    hits = response['hits']['hits']
    scroll_id = response["_scroll_id"] #获取第三次scroll_id