7 Elasticsearch Basic CRUD Operation Examples – Index, Update, Get and Delete

by Santosh Yadav on May 22, 2019

[Elasticsearch CURD Operations] Elasticsearch supports storing documents in JSON format. It also provides REST interface to interact with elasticsearch datastore.

In this article, we will discuss how to do basic CRUD operations on elasticsearch datastore using the following examples:

1. Index API – Index a document by providing document id
2. Index API – Index a document with auto-generated document id
3. Get API – Retrieve a document along with all fields
4. Get API – Retrieve a document along with specific fields
5. Delete API – Delete a document from datastore
6. Update API – Update the whole document
7. Update API – Update only partial document (adding new fields)

These operation fall under document APIs, they are named so because they deal with documents. The good thing about elasticsearch is that we don’t need to create database schema beforehand. We could start inserting data even before creating database schema.

Schema equivalent in these context is mapping. We can create a mapping that is quite similar to JSON schema of documents we want to insert. In current context if we don’t provide mapping of documents, elasticsearch will infer at run time the data type of all the elements of JSON document.

Database equivalent in this context is INDEX, Table equivalent in this context is TYPE.

First, we need to create an index that we can use in our examples.

Insertion of documents in elasticsearch is called indexing of documents. So let’s first create our articles database i.e. articles index

$ curl -XPUT '192.168.101.100:9200/articles?&pretty'

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100    84  100    84    0     0     84      0  0:00:01 --:--:--  0:00:01   103{
"acknowledged" : true,
"shards_acknowledged" : true,
"index" : "articles"
}

A simple PUT command will create the articles index, now we can index our article documents within this index. In command response we can see index is created.

When we are indexing a document, elasticsearch assigns a document id to each indexed document, though we have the option of choosing this document id, we can leave it to elasticsearch. Using this id we can retrieve JSON documents later.

The following is ouor sample JSON document that we will use for the rest of the examples in this tutorial.

{
"topic": "python",
"title": "python tuples",
"description": "practical operations with python tuples",
"author": "santosh",
"date": "1-1-2019",
"views": "100"
}

Note: If you are new to elasticsearch, this will get you started: How to Install and Configure Elasticsearch on Linux and Windows

1. Index API – Index a document by providing document id

In this example we will index the document within _doc type of articles index, with document id 1

curl -XPOST '192.168.101.100:9200/articles/_doc/1?pretty' -d '{"topic":"python","title": "python tuples","description": "practical operations with python tuples","author": "santosh","date": "1-1-2019","views" : "100"}' -H 'Content-Type: application/json'

The following is the output of the above command.

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   377  100   222  100   155    222    155  0:00:01 --:--:--  0:00:01  1008{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 1,
"result" : "created",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 0,
"_primary_term" : 1
}

After ip and port, next is index then type within index then document id

Document is indexed with document id 1.

2. Index API – Index a document with auto-generated document id

If we don’t provide document while indexing, elasticsearch will auto generate the document id for indexed documents.

curl -XPOST '192.168.101.100:9200/articles/_doc/?pretty' -d '{"topic":"python","title": "python tuples","description": "practical operations with python tuples","author": "santosh","date": "1-1-2019","views" : "100"}' -H 'Content-Type: application/json'

The following is the output of the above command.

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   396  100   241  100   155    241    155  0:00:01 --:--:--  0:00:01  2117{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1zfK-2kBx40Oa0-N-vjk",
"_version" : 1,
"result" : "created",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 0,
"_primary_term" : 1

After ip and port, next is index then type within index. In this example we won’t provide document id.

In the response, elasticsearch provides the document id “_id” : “1zfK-2kBx40Oa0-N-vjk”,

3. Get API – Retrieve a document along with all fields

Using Get API we can retrieve documents from elasticsearch datastore. Documents are retrieved using document id, let’s retrieve document with id 1

curl -XGET '192.168.101.100:9200/articles/_doc/1?pretty'

Output:

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   368  100   368    0     0    368      0  0:00:01 --:--:--  0:00:01  3956{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 1,
"_seq_no" : 0,
"_primary_term" : 1,
"found" : true,
"_source" : {
"topic" : "python",
"title" : "python tuples",
"description" : "practical operations with python tuples",
"author" : "santosh",
"date" : "1-1-2019",
"views" : "100"
}
}

As you can see retrieved documents.

Let’s try to retrieve document which does not exist

$ curl -XGET '192.168.101.100:9200/articles/_doc/11?pretty'

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100    83  100    83    0     0     83      0  0:00:01 --:--:--  0:00:01   761{
"_index" : "articles",
"_type" : "_doc",
"_id" : "11",
"found" : false
}

Response shows found=false, which means the document does not exist.

4. Get API – Retrieve a document along with specific fields

In next example we will do selective GET , I.e. we will request only certain elements from elasticsearch datastore

curl -XGET '192.168.101.100:9200/articles/_doc/1?pretty&_source=topic,title,author'

Output:

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   240  100   240    0     0    240      0  0:00:01 --:--:--  0:00:01  2580{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 3,
"_seq_no" : 2,
"_primary_term" : 1,
"found" : true,
"_source" : {
"author" : "santosh",
"topic" : "python",
"title" : "python tuples"
}
}

5. Delete API – Delete a document from datastore

Using Delete API we can delete a document from elasticsearch datastore, to delete a document we need document id

curl -XDELETE '192.168.101.100:9200/articles/_doc/1zfK-2kBx40Oa0-N-vjk?pretty'

Output:

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   241  100   241    0     0    241      0  0:00:01 --:--:--  0:00:01  1928{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1zfK-2kBx40Oa0-N-vjk",
"_version" : 2,
"result" : "deleted",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 1,
"_primary_term" : 1
}

In response the result is deleted, that means document is deleted.

Let’s try to delete a document that does not exist

$ curl -XDELETE '192.168.101.100:9200/articles/_doc/1zfK-2kBx40Oa0-N-vjk?pretty'

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   243  100   243    0     0    243      0  0:00:01 --:--:--  0:00:01  3115{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1zfK-2kBx40Oa0-N-vjk",
"_version" : 3,
"result" : "not_found",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 2,
"_primary_term" : 1
}

In response we can see, result is not found.

6. Update API – Update the whole document

Using this API we can update existing document stored in elasticsearch datastore. To update the document we need document id of the document.

Let’s update the whole document or replace an existing document. We will replace the document stored with document id 1

curl -XPOST '192.168.101.100:9200/articles/_doc/1?pretty' -d '{"topic":"python","title": "python tuples","description": "practical operations with python sets","author": "santosh","date": "11-11-2019","views" : "1000"}' -H 'Content-Type: application/json'

Output:

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   378  100   222  100   156    222    156  0:00:01 --:--:--  0:00:01  3024{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 2,
"result" : "updated",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 1,
"_primary_term" : 1
}

View the updated document:

$ curl -XGET '192.168.101.100:9200/articles/_doc/1?pretty'

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   350  100   350    0     0    350      0  0:00:01 --:--:--  0:00:01  4487{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 2,
"_seq_no" : 1,
"_primary_term" : 1,
"found" : true,
"_source" : {
"topic" : "python",
"title" : "python tuples",
"description" : "practical operations with python sets",
"author" : "santosh",
"date" : "11-11-2019",
"views" : "1000"
}
}

As you can see document is replaced with new document , and version is incremented to 2 , incrementing version represents that this document has been modified .

7. Update API – Update only partial document (adding new fields)

In this example we will add two new fields in existing document.

curl -XPOST '192.168.101.100:9200/articles/_doc/1/_update?pretty' -d '{"doc":{"uniqueviews":"1789","reviewer":"RN"}}' -H 'Content-Type: application/json'

Output:

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   268  100   222  100    46    222     46  0:00:01 --:--:--  0:00:01  1914{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 3,
"result" : "updated",
"_shards" : {
"total" : 2,
"successful" : 2,
"failed" : 0
},
"_seq_no" : 2,
"_primary_term" : 1
}

View the updated document:

$ curl -XGET '192.168.101.100:9200/articles/_doc/1?pretty'

% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
            Dload  Upload   Total   Spent    Left  Speed
100   401  100   401    0     0    401      0  0:00:01 --:--:--  0:00:01  5141{
"_index" : "articles",
"_type" : "_doc",
"_id" : "1",
"_version" : 3,
"_seq_no" : 2,
"_primary_term" : 1,
"found" : true,
"_source" : {
"topic" : "python",
"title" : "python tuples",
"description" : "practical operations with python sets",
"author" : "santosh",
"date" : "11-11-2019",
"views" : "1000",
"uniqueviews" : "1789",
"reviewer" : "RN"
}
}

In the above GET command output, we can see the two new fields (uniqueviews and reviewer) are added to the document.

Add your comment