Get started with Hyper-Converged Database (HCD)
Hyper-Converged Database (HCD) is a self-managed database built on Apache Cassandra® that empowers you to manage your data infrastructure with enterprise-grade capabilities. This topic provides essential information for users new to HCD.
HCD provides all the capabilities of Apache Cassandra®, as well as support for vector search.
HCD is ideal for enterprise operators and architects who want to take advantage of datacenter modernization investments to reduce operating costs and operational overhead. HCD supports modern datacenters and HCI for data workloads and AI systems at scale.
|
To try a limited non-production HCD deployment, you can run an HCD Docker container. For more information, see Install Hyper-Converged Database (HCD) using Docker for non-production local containerized environments. |
Operations platform
Mission Control is the operations platform for HCD. It simplifies management of all HCD operations across an array of hosting options from the cloud to on-premises and everything in-between.
Mission Control focuses on the lifecycle management, security, operations, and observability of HCD clusters. It handles the orchestration of automation across regional cluster boundaries. This centralizes management of globally deployed clusters in a single location.
DataStax strongly recommends using Mission Control to provision and manage your clusters. To get started with Mission Control, see Mission Control installation requirements.
Install HCD
Self-managed clusters require planning to ensure that you have the appropriate hardware, software, and configurations to support your workloads. To get started planning your cluster, see Capacity planning and hardware selection for HCD deployments.
After you deploy a cluster, install HCD on the nodes in the cluster.
There are several ways to deploy HCD clusters. Select the method that best suits your environment:
-
Mission Control (recommended):
Mission Control is a cloud-based service that provides a unified management console for your database clusters. If you are planning a production deployment, DataStax recommends using Mission Control to manage your clusters.
Use the Mission Control UI to install HCD without running Kubernetes manually. The UI provides simple and expert modes.
If you have Kubernetes experience, you can use the CLI to run
kubectlcommands for installation.Mission Control also deploys the Data API, a JSON document oriented interface to your data.
-
Install HCD on any supported Linux-based platform. Doesn’t require root permissions.
-
Installs HCD on Debian-based systems using the APT package manager. Runs HCD as a service.
-
Installs HCD on RHEL-based systems using the Yum package manager. Runs HCD as a service.
-
Docker container (demo only):
Use Docker to explore non-production development with HCD and the Data API in a local containerized environment.
Develop applications with HCD
HCD is designed for vector search applications, such as generative AI (gen AI), semantic search, and geospatial search, as well as traditional schema-driven NoSQL applications.
HCD stores vector embeddings alongside non-vector data. If you build Retrieval Augmented Generation (RAG)-based applications, such as assistants and chatbots, you can use HCD to store and query vector embeddings.
For applications that require fixed-schema and schemaless data, you can store both types of data in HCD, and then query the appropriate dataset through the corresponding API.
Connect to HCD
HCD 1.2 provides a variety of APIs for developing applications:
-
Data API with clients: The Data API is DataStax’s API for writing applications that store and query unstructured document data and fixed schema CQL tables. DataStax provides several clients to facilitate interaction with the Data API.
-
CQL API with drivers: The Cassandra Query Language (CQL) is a SQL-like language for querying and managing databases. It stores structured data in tables and uses primary keys to index data. The main development tool for writing applications that use CQL are Cassandra drivers. For more information, see Cassandra driver compatibility.
-
CQL API with
cqlsh: The CQL shell (cqlsh) is a command-line utility for executing CQL commands. If you are in the development phase and want to quickly test queries,cqlshis a good tool to use. See instructions for usingcqlshin Connect to HCD with the CQL shell (cqlsh). The CQL documentation provides information on all aspects of creating schema and querying data.
For details, see Connect to Hyper-Converged Database (HCD).
Load data
Use these tools to load data into your databases:
-
Streaming tools, such as the DataStax Apache Kafka Connector
