​
Engine Tiers

We have different engine tiers to match your compute and storage needs. Each tier provides different levels of resources and capabilities.
  • Sandbox: 1CU shared engine for development and testing, no permanent storage.
  • Boost: Shared engine that can boost up to 8 CUs, great for applications with varying performance requirements.
  • XS (Extra Small): 1 CU dedicated engine with unlimited storage, ideal for small production workloads.
  • S (Small): 2 CU dedicated engine with unlimited storage, perfect for growing applications and small team analytics.
  • M (Medium): 4 CU dedicated engine with unlimited storage, designed for production applications and moderate data processing.
  • L (Large): 8 CU dedicated engine with unlimited storage, built for high-traffic applications and intensive analytics.

​
Compute Units (CU)

A Compute Unit (CU) represents roughly:
  • 1 dedicated vCPU
  • 8GB of RAM
  • 100GB of fast local NVMe storage space (used for caching and query disk spill)

​
Storage

All engine tiers (except Sandbox) include:
  • Unlimited storage backed by object storage
  • Fast local cache for frequently accessed data
  • Automatic scaling - storage grows as needed
  • High durability with built-in redundancy
The local NVMe cache provides high-speed access to your most frequently queried data, while the object storage backend ensures unlimited capacity and data durability.

​
Managing Engines

You can create, start, stop, and manage engines using the brightstack CLI: 
# Create an engine
bright engine create --name <ENGINE_NAME> --type <ENGINE_TYPE>

# List your engines
bright engine list

# Start an engine
bright engine start <ENGINE_ID_OR_NAME>

# Stop an engine
bright engine stop <ENGINE_ID_OR_NAME>
Engines that have been idle for over 10 minutes will be automatically stopped to save costs. You can start them back up anytime you need them.