Skip to main content

OpenStack Sahara

OpenStack Sahara is an OpenStack project that provisions and manages data processing clusters, such as Apache Hadoop and Apache Spark, on OpenStack infrastructure for on-demand big data workloads (data processing / big data infrastructure).

  • On-demand provisioning and lifecycle management of Hadoop and Spark clusters on OpenStack (data processing orchestration).
  • Cluster templates, node group templates, and image management for repeatable big data environments (configuration management).
  • Integration with core OpenStack services such as Compute (Nova), Networking (Neutron), and Object Storage (Swift) (cloud infrastructure).
  • Support for multiple data processing frameworks and plugins, including Hadoop and Spark distributions (big data platforms).
  • Representational State Transfer (REST) Application Programming Interface (API) and dashboard integration with Horizon for programmatic and UI-based cluster operations (API and cloud management UI).

More About OpenStack Sahara

OpenStack Sahara is a data processing service in the OpenStack ecosystem that automates the deployment and management of big data frameworks, including Apache Hadoop and Apache Spark, on OpenStack-based clouds (data processing / big data infrastructure). It addresses the need for users to create and operate transient or persistent data processing clusters without manually configuring virtual machines, networking, and storage for each workload.

Sahara focuses on orchestrating clusters of virtual machines that run big data engines (cluster orchestration). It uses concepts such as cluster templates and node group templates (configuration management) to describe cluster topologies, roles, and scaling characteristics. These templates enable operators and data teams to standardize cluster layouts, such as master, worker, and edge nodes, and to reuse them across multiple deployments.

The service integrates with core OpenStack components, including Compute (Nova) for Virtual Machine (VM) provisioning, Networking (Neutron) for network configuration, and Object Storage (Swift) and Block Storage (Cinder) for data storage (cloud infrastructure). Sahara also plugs into the OpenStack Identity service (Keystone) for authentication and multi-tenant access control (identity and access management). This alignment with OpenStack services allows Sahara-deployed clusters to fit into existing OpenStack operational, security, and quota models.

Sahara supports multiple data processing frameworks through a plugin architecture (extensibility framework). Plugins encapsulate knowledge about specific Hadoop and Spark distributions and their configuration. This enables operators to choose frameworks and distributions suitable for their environment while using a consistent provisioning workflow. Sahara also works with image management via OpenStack Image service (Glance), where images are prepared with the required big data software and referenced by cluster templates.

For access and control, Sahara exposes a RESTful API (API / integration) and integrates with the OpenStack Horizon dashboard (cloud management UI). Through Horizon, users can define templates, launch clusters, monitor their status, and scale clusters as needed. The REST API supports automation, integration with Continuous Integration and Continuous Deployment (CI/CD) pipelines, and programmatic scheduling of transient clusters for specific analytical or batch processing jobs.

In enterprise and institutional environments, Sahara is used to provide big data processing capabilities as a service on private or public OpenStack clouds. It enables data engineers and analysts to obtain Hadoop or Spark clusters on demand, aligned with cloud governance, quotas, and security policies. Within a technical directory or taxonomy, OpenStack Sahara fits in categories such as data processing orchestration, big data on cloud infrastructure, and OpenStack ecosystem services.