Technical Reference Deep Dives¶

Internals¶

The following is a dive into some of the internals in nova.

• AMQP and Nova: How nova uses AMQP as an RPC transport
• Scheduling: The workflow through the scheduling process
• Live Migration: The live migration flow
• Services, Managers and Drivers: Module descriptions for some of the key modules used in starting / running services
• Virtual Machine States and Transitions: Cheat sheet for understanding the life cycle of compute instances
• Threading model: The concurrency model used in nova, which is based on eventlet, and may not be familiar to everyone.
• Notifications in Nova: How the notifications subsystem works in nova, and considerations when adding notifications.
• ComputeDriver.update_provider_tree: A detailed explanation of the ComputeDriver.update_provider_tree method.

Debugging¶

• Guru Meditation Reports: Inspired by Amiga, a way to trigger a very comprehensive dump of a running service for deep debugging.

Forward Looking Plans¶

The following section includes documents that describe the overall plan behind groups of nova-specs. Most of these cover items relating to the evolution of various parts of nova’s architecture. Once the work is complete, these documents will move into the “Internals” section.

If you want to get involved in shaping the future of nova’s architecture, these are a great place to start reading up on the current plans.

• Cells: Comparison of Cells v1 and v2, and how v2 is evolving
• Rest API Policy Enforcement: How we want policy checks on API actions to work in the future
• Nova Stable REST API: What stable api means to nova
• Scheduler Evolution: Motivation behind the scheduler / placement evolution