Why choose CrateDB drivers?
- Language-agnostic connectivity: Whether you use Java, Python, .NET, Node.js or Go, CrateDB provides first-class drivers and libraries that speak your language.
- Full SQL support: These libraries connect via the PostgreSQL Wire Protocol or HTTP/REST endpoint, enabling you to leverage CrateDB’s full power (real-time analytics, distributed joins, search, vector operations) without limiting yourself to a narrow interface.
- Developer-friendly: Easy installation, familiar APIs, and documentation tailored for each environment, so you spend less time integrating and more time building.
- Seamless scalability: Because CrateDB is designed for real-time, large-scale analytics on structured, semi-structured and unstructured data, the drivers are built to perform under scale, letting you focus on analytics, not plumbing.
What drivers are available?
Here’s a quick overview of popular drivers and their key features:
| Language / Platform | Driver / Library | Highlights |
|---|---|---|
| Java | JDBC driver | Based on PostgreSQL JDBC, pure Java, enables use via standard java.sql ecosystem. |
| Python | crate client library |
Implements Python DB API 2.0 spec; install via pip install crate. |
| Node.js | node-crate (or REST interface) |
Connects via HTTP/REST _sql endpoint; ideal for JavaScript/TypeScript applications. |
| Go, .NET, PHP, … | Adapters & connectors | Available via ecosystem (via PostgreSQL-wire or HTTP/REST) as noted in the “Drivers & Tools” reference. |
Getting started: example (Python)
pip install crate
from crate import client
connection = client.connect("http://localhost:4200", username="user", password="pwd")
cursor = connection.cursor()
cursor.execute("SELECT col1, count(*) FROM my_table GROUP BY col1")
for row in cursor:
print(row)
Best practices for drivers & clients
-
Use the correct protocol: For high-performance workloads you might prefer the PostgreSQL Wire Protocol drivers (e.g., JDBC, standard PostgreSQL clients). For flexible HTTP/REST access (e.g., serverless functions, scripts, polyglot environments) you can use the HTTP endpoint + clients.
-
Configure connection pooling wisely: Especially in production environments with many concurrent queries, ensure your client library supports pooling or reuse of connections for low latency.
-
Leverage async or batch features: Where supported (especially in Go, Node.js, Python), use asynchronous query execution or bulk insert/updates to handle high-throughput ingestion or export workflows.
-
Monitor driver health and compatibility: Keep your client library versions aligned with the CrateDB server version; monitor for driver updates and compatibility releases.
-
Secure your connections: Always use TLS/HTTPS or encrypted wire protocol when connecting across networks; manage credentials, roles and tokens with care.
When to use which client / driver
-
Use a native driver (e.g., JDBC for Java, Python
cratelib) when you need full-featured access, complex query support, connection pooling and integration into complex applications. -
Use the HTTP/REST interface (via generic HTTP library or lightweight client) when you need language-agnostic access, microservices, event triggers, or simple query layers.
-
Use ODBC or generic PostgreSQL clients if you have existing tools (BI dashboards, ETL frameworks) built against PostgreSQL-compatible drivers—you can point them to CrateDB thanks to its PostgreSQL wire compatibility
