Elastic search

Elasticsearch is a distributed search and analitycs engine. You can add documents, and perform analitycs over them.

For more details check:

• Try Elasticsearch and Kibana: for guide on local deployment using docker or k8s.

• Elasticsearch labs: for projects that use Elasticsearch and notebooks with code snippets.

Connection

Elasticsearch is designed to be run as a separate service. For local installation, check out try Elasticsesarch and Kibana.

Note: don’t forget to set: ELASTICSEARCH_USER=elastic and ELASTICSEARCH_PASSWORD=elastic environment variables before starting the service.

An instance of the Elasticsearch class represents a connection Elasticsearch.

---

The following cell shows the connection to elasticsearch.

from elasticsearch import Elasticsearch
es = Elasticsearch(
    hosts=["http://localhost:9200"],
    basic_auth=("elastic", "elastic")
)
es

<Elasticsearch(['http://localhost:9200'])>

Use the info method to unsure the connection is established correctly. The following cell shows possible outputs of the info method.

from pprint import pprint
pprint(dict(es.info()))

{'cluster_name': 'docker-cluster',
 'cluster_uuid': 'BDjLMAgzRxa9TP634NQHew',
 'name': '5a0053e0a328',
 'tagline': 'You Know, for Search',
 'version': {'build_date': '2025-08-24T22:05:04.526302670Z',
             'build_flavor': 'default',
             'build_hash': '0c781091a2f57de895a73a1391ff8426c0153c8d',
             'build_snapshot': False,
             'build_type': 'docker',
             'lucene_version': '10.2.2',
             'minimum_index_compatibility_version': '8.0.0',
             'minimum_wire_compatibility_version': '8.19.0',
             'number': '9.1.3'}}

Index

In the context of Elasticsearch, an index is a collection of similar documents. Analog in relational databases is “table”.

Action	Python Method	Description
Create an Index	es.indices.create()	Creates a new index. You can pass settings and mappings as keyword arguments to define the index structure.
Check for an Index	es.indices.exists()	Returns True if an index with the given name exists, otherwise returns False.
Get Index Information	es.indices.get()	Retrieves detailed information about an index, including its settings and mappings.
Delete an Index	es.indices.delete()	Permanently deletes an index and all of its data.

---

Consider the basic operations of indices.

The following cell creates the “book_index”. mappings specify the attributes index contains and their types.

mappings = {
    "properties": {
        "prop1": {"type": "text"},
        "prop2": {"type": "keyword"}
    }
}

es.indices.create(index="book_index", mappings=mappings)

ObjectApiResponse({'acknowledged': True, 'shards_acknowledged': True, 'index': 'book_index'})

Now, the method es.indices.exists returns True as index has been created.

es.indices.exists(index="book_index")

HeadApiResponse(True)

The information provided by the get method is represented above.

pprint(dict(es.indices.get(index="book_index")))

{'book_index': {'aliases': {},
                'mappings': {'properties': {'prop1': {'type': 'text'},
                                            'prop2': {'type': 'keyword'}}},
                'settings': {'index': {'creation_date': '1756809943562',
                                       'number_of_replicas': '1',
                                       'number_of_shards': '1',
                                       'provided_name': 'book_index',
                                       'routing': {'allocation': {'include': {'_tier_preference': 'data_content'}}},
                                       'uuid': 'G9uweKCnSFyU_aHWSsI8Zw',
                                       'version': {'created': '9033000'}}}}}

As expected, the exists method returns False after the index is deleted.

es.indices.delete(index="book_index")
es.indices.exists(index="book_index")

HeadApiResponse(False)