Jump to: navigation, search

Difference between revisions of "Compass"

(Why Compass?)
(Why Compass?)
Line 17: Line 17:
  
 
* Small Code Base:  
 
* Small Code Base:  
** Currently, the total code base is around 5,000 line of Python code
+
** The total code base is around 5,000 line of Python code
  
 
= Compass Overview =
 
= Compass Overview =
  
 
= Compass Use Cases =
 
= Compass Use Cases =

Revision as of 00:30, 1 November 2013

What is Compass?

Compass is an open source project trying to providing "deployment as a service" to a set of bare metal machines.

It can provide a Restful API service for operators to pragmatically deploy an OpenStack system on to the raw physical resources, from 10,000 feet view, it can be considered as a Live CD to a set of physical machines to the end users (staff deploying OpenStack system).

Why Compass?

  • Deployment System Extensibility: Compass is not limited to OpenStack cluster, it provides a data-driven framework for deployment.
    • Example: Providing capability to create a Ceph cluster by extending the data fields in Compass system
  • Hardware Support Extensibility: Like the Cinder's supports multiple storage vendors' system controlled by OpenStack API through plug-in framework, Compass supports topology-aware resource discovery to multiple networking vendor's equipment through its plug-in framework.
    • Example: Compass originally supports Huawei switches for topology-aware resource discovery, and we added support to HP switches with less than a couple of hundred lines of plug-in code.
  • Host OS/Hypervisor Extensibility:
    • Example: we first streamlined the process for CentOS, and Ubuntu support is being supported at this moment. We imagine we can repeat the similar easy effort to systems like ESXi.
  • Small Code Base:
    • The total code base is around 5,000 line of Python code

Compass Overview

Compass Use Cases