一、Reindex重建索引
1 POST _reindex
2 {
3 "source": {
4 "index": "ecgdata"
5 },
6 "dest": {
7 "index": "ecg"
8 }
9 }
二、index template
1 #创建索引模板my_index_template定义mappings
2 #要求：
3 #只要名字以index_开头的index，都用这个mappings
4 #设置主分片3，副本分片1
5 #int_开头的字段都是用integer类型
6 #字符串类型的字段都使用keyword并且使用english分词器
7 PUT _index_template/template_1
8 {
9 "index_patterns": [
10 "index_*",
11 "bar*"
12 ],
13 "template": {
14 "settings": {
15 "number_of_shards": 3,
16 "number_of_replicas": 1
17 },
18 "mappings": {
19 "dynamic_templates": [
20 {
21 "integers": {
22 "match_mapping_type": "string",
23 "mapping": {
24 "type": "keyword",
25 "analyzer": "english"
26 }
27 }
28 },
29 {
30 "longs_as_strings": {
31 "match": "int_*",
32 "mapping": {
33 "type": "integer"
34 }
35 }
36 }
37 ]
38 }
39 }
40 }
三、Component template
组件模板，相当于更细粒度的索引模板。一个索引模板可以由多个组件模板组成
1 PUT _component_template/mappings
2 {
3 "template": {
4 "settings": {
5 "number_of_shards": 3,
6 "number_of_replicas": 1
7 }
8 }
9 }
10
11 PUT _index_template/tem
12 {
13 "index_patterns": ["te*","bar*"],
14 "priority":500,
15 "composed_of":["mappings"]
16 }
四、Update_by_query
是在原数据上进行操作，若考试的时候不确定是否操作正确则先reindex，在新索引上update_by_query结果正确再到考试索引中操作
1 POST ecgdata/_update_by_query
2 {
3 "script": {
4 "source": "ctx._source.price-=1000",
5 "lang": "painless"
6 },
7 "query": {
8 "term": {
9 "price": "4999"
10 }
11 }
12 }
五、Pipeline
考试中如果要使用pipeline，就要在elasticsearch.yml文件中配置ingest角色(预处理)
1、pipeline+reindex
1 #索引index_pipeline中,有一个数组对象tags,各个数组的元素中,有些前面又空格,有些后面有空格,通过定义ingest pipeline,将这些数组元素中的空格给去掉。另外要求通过reindex将index_pipeline这个索引,应用于刚刚定义好的pipeline,从而生成一个新的索引index_pipeline_new
2 POST index_pipeline/_bulk
3 {"index":{"_id":1}}
4 {"tags":["ping pang", "basket ball", " foot ball "]}
5 {"index":{"_id":2}}
6 {"tags":[" ping pang ", "gof ball"]}
7
8 PUT _ingest/pipeline/my-pipeline
9 {
10 "processors": [
11 {
12 "foreach": {
13 "field": "tags",
14 "processor": {
15 "trim": {
16 "field": "_ingest._value"
17 }
18 }
19 }
20 }
21 ]
22 }
23
24 POST _reindex
25 {
26 "source": {
27 "index": "index_pipeline"
28 },
29 "dest": {
30 "index": "index_pipeline_new",
31 "pipeline": "my-pipeline"
32 }
33 }
2、pipeline+update_by_query
1 #现有索引task2中，每个文档都有value01、value02、value03这三个字段。为这个索引中的所有文档，新增一个字段，名称为newadd，这个字段的值，是value01、value02、value03这三个字段的和(拼接)
2 PUT task2/_doc/1
3 {
4 "value01":"asd",
5 "value02":"sdg",
6 "value03":"wer"
7 }
8 PUT task2/_doc/2
9 {
10 "value01":"asdrr",
11 "value02":"sdgrr",
12 "value03":"werrr"
13 }
14
15 PUT _ingest/pipeline/my-pipeline
16 {
17 "processors": [
18 {
19 "set": {
20 "field": "value04",
21 "value": "{{{value01}}}{{{value02}}}{{{value03}}}"
22 }
23 }
24 ]
25 }
26
27 POST task2/_update_by_query?pipeline=my-pipeline
六、Enrich processor
按照官网的顺序直接cv
1 #source index
2 PUT /users/_doc/1?refresh=wait_for
3 {
4 "email": "mardy.brown@asciidocsmith.com",
5 "first_name": "Mardy",
6 "last_name": "Brown",
7 "city": "New Orleans",
8 "county": "Orleans",
9 "state": "LA",
10 "zip": 70116,
11 "web": "mardy.asciidocsmith.com"
12 }
13
14 #创建 enrich 策略
15 PUT /_enrich/policy/users-policy
16 {
17 "match": {
18 "indices": "users", #source index的名称
19 "match_field": "email", #哪个字段与source index进行关联
20 "enrich_fields": ["first_name", "last_name", "city", "zip", "state"] #将哪些字段放到target索引中，这些字段来源与source index
21 }
22 }
23
24 #执行 enrich策略
25 POST /_enrich/policy/users-policy/_execute
26
27 #定义pipeline 使用enrich策略
28 PUT /_ingest/pipeline/user_lookup
29 {
30 "processors" : [
31 {
32 "enrich" : {
33 "description": "Add 'user' data based on 'email'",
34 "policy_name": "users-policy",
35 "field" : "email", #target index中的字段
36 "target_field": "user", #要创建字段的名字
37 "max_matches": "1"
38 }
39 }
40 ]
41 }
42
43 #创建索引，使用pipeline
44 PUT /my-index-000001/_doc/my_id?pipeline=user_lookup
45 {
46 "email": "mardy.brown@asciidocsmith.com"
47 }