如下所示:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
|
""" 提取文档数超过10000的数据 按照某个字段的值具有唯一性进行升序, 按照@timestamp进行降序, 第一次查询,先将10000条数据取出, 取出最后一个时间戳, 在第二次查询中,设定@timestamp小于将第一次得到的最后一个时间戳, 同时设定某个字段的值具有唯一性进行升序, 按照@timestamp进行降序, """ from elasticsearch import Elasticsearch import os write_path = "E:\\公司\\案例数据采集\\olt告警案例分析\\10000_data.txt" es = Elasticsearch(hosts = "", timeout = 1500 ) write_file = open (write_path, "a+" ) def _first_query(): index_ = "gather-010" _source = [ "TWICE_BOOK_TIME" , "@timestamp" ] try : rs = es.search(index = index_, body = { "size" : 10000 , "query" : { "match_all" : {} }, "sort" : [ { "@timestamp" : { "order" : "desc" } }, { "TASK_RECEIVE_ID.keyword" : { "order" : "asc" } } ], "_source" : _source }) return rs except : raise Exception( "{0} search error" . format (index_)) def _get_first_data(first_rs): i = 0 if first_rs: for hit in first_rs[ 'hits' ][ 'hits' ]: IptvAccount = hit[ '_source' ][ 'TWICE_BOOK_TIME' ] timestamp = hit[ '_source' ][ '@timestamp' ] if IptvAccount is None : IptvAccount = "" write_file.write(IptvAccount + "," + timestamp + "\n" ) i + = 1 if i = = 10000 : return timestamp def _second_query(timestamp): index_ = "gather-010" _source = [ "TWICE_BOOK_TIME" , "@timestamp" ] try : rs = es.search(index = index_, body = { "size" : 10000 , "query" : { "bool" : { "filter" : { "range" : { "@timestamp" : { "lt" : timestamp } } } } }, "sort" : [ { "@timestamp" : { "order" : "desc" } }, { "TASK_RECEIVE_ID.keyword" : { "order" : "asc" } } ], "_source" : _source }) return rs except : raise Exception( "{0} search error" . format (index_)) if __name__ = = "__main__" : first_rs = _first_query() first_timestamp = _get_first_data(first_rs) print (first_timestamp) while True : second_rs = _second_query(first_timestamp) first_timestamp = _get_first_data(second_rs) if first_timestamp is None : break print (first_timestamp) |
以上这篇python批量从es取数据的方法(文档数超过10000)就是小编分享给大家的全部内容了,希望能给大家一个参考,也希望大家多多支持服务器之家。
原文链接:https://blog.csdn.net/sxf_123456/article/details/80373818