2.0 KiB
OPENSTACK DIAGNOSTICS PROPOSAL
Overview
OpenStack cloud operators usually rely on deploymnet tools to configure all the platform components correctly and efficiently upfront. However, after initial deployment platform configurations and operational conditions start to change. We propose a project that allows to analyze OpenStack architecture and diagnose existing and potential problems using flexible set of rules.
Mission
Diagnostics project mission is to provide OpenStack cloud operators with flexible way to inspect, analyze and diagnose architecture of the cloud and configuration of components of the platform.
User Stories
As a cloud operator, I want to be able to automatically extract configuration parameters from all OpenStack components to verify their correctness, consistency and integrity. As a cloud architect, I want to make sure that my OpenStack architecture and configuration are compliant to 'best practices'. As a cloud operator, I want automatic diagnostics tool which can tell me what problems does my OpenStack architecture and/or configuration have or might potentially have (e.g. at scale or if some component or node failed). As a cloud operator, I want to be able to define rules used to inspect and verify configuration of OpenStack components and store them to use for verification of future configuration changes.
Requirements
TBD
Scope
As an MVP1, we create service that includes:
- Rules engine with grammatic analysis capabilities
- Extensible implementation of rules
- REST API for running inspections
- Storage back-end implementation for OpenStack platform architecture and configuration data
Assumptions
We assume that we must reuse as much as possible from OpenStack Deployment program in terms of platform configuration and architecture definitions (i.e. TripleO Heat and configuration files templates).