_connection-9.3-python.md

endpoint	_connection
lang	python
es_version	9.3
client	elasticsearch==9.3.0

Elasticsearch 9.3 connection management (Python example)

Use the Elasticsearch class from the elasticsearch package to create a client. The simplest approach uses basic authentication with a username and password:

from elasticsearch import Elasticsearch

client = Elasticsearch(
    "http://localhost:9200",
    basic_auth=("elastic", "your-password"),
)

info = client.info()
print(f"Connected to {info['cluster_name']} (v{info['version']['number']})")

Install the client with pip:

pip install elasticsearch==9.3.0

API key authentication

For production use, API keys are preferred over username/password. Pass the base64-encoded API key directly:

client = Elasticsearch(
    "http://localhost:9200",
    api_key="your-base64-api-key",
)

Or use a tuple of (id, api_key) if you have them separately:

client = Elasticsearch(
    "http://localhost:9200",
    api_key=("key-id", "key-secret"),
)

Connection options

Configure timeouts, retries, and TLS for production deployments:

client = Elasticsearch(
    "https://my-cluster.example.com:9243",
    api_key="your-base64-api-key",
    request_timeout=30,
    max_retries=3,
    retry_on_timeout=True,
    ca_certs="/path/to/ca.crt",
)

To verify the connection is working:

if client.ping():
    print("Connected")
else:
    print("Connection failed")