[Starlingx-discuss] RFC: Project Priorities

Jones, Bruce E bruce.e.jones at intel.com
Fri Jul 13 05:03:22 UTC 2018


Thank you Brent, your changes look good.

       brucej

-----Original Message-----
From: Rowsell, Brent [mailto:Brent.Rowsell at windriver.com] 
Sent: Thursday, July 12, 2018 2:23 PM
To: Jones, Bruce E <bruce.e.jones at intel.com>; starlingx-discuss at lists.starlingx.io
Subject: RE: RFC: Project Priorities

Bruce,

I have added some initial comments and new items to the etherpad. More to come 

Brent 

-----Original Message-----
From: Jones, Bruce E [mailto:bruce.e.jones at intel.com] 
Sent: Thursday, July 12, 2018 2:18 PM
To: starlingx-discuss at lists.starlingx.io
Subject: [Starlingx-discuss] RFC: Project Priorities

We had an internal meeting yesterday to discuss project priorities.  I'd like to share our list with the community, get feedback, and submit this as input to the Architect call next Thursday.

This list is also available on https://etherpad.openstack.org/p/stx-planning, which I propose we use as the master for this discussion.

              brucej

Project Priorities
=============

1. Build stability & mirror cache / improvements 2. Basic processes & infrastructure  e.g. CI/CD, Test/Validation, Releases, Dev, Bug handling, Security, Devstack, Zuul tests, unit tests, etc....
3. Documentation  e.g. project map, how to contribute, how to build/install, how to run, how to debug, ....
4. Distro management - build, updates, automation, etc...

We believe these four items are needed for StarlingX to be a viable project and to recruit & retain contributors, users and operators.  

5. KPI establishment, test automation and on-going improvements

The value of StarlingX is in the high level of performance and robustness it provides.  Setting KPIs and improving them over time is critical.

6. Reduce patch backlog - openstack and non-openstack

The patch backlog is something no one likes.  The Intel team has the lead on this but on-going assistance from Wind River will be needed (e.g. to help write upstream Specs).  There are many opportunities to re-architect the system to more closely align it with upstream communities.

7. Containerized OpenStack services and support for containerized workloads

We believe this is a key feature for the product.  Wind River has the lead on this work.

8. Networking features (containerized VNFs, ONAP integration, TSN, ...)

High performance and fully featured networking is another key value of StarlingX.  Adding containerized VNFs, supporting ONAP and Time Senstive Networking enable important use cases for the project.

9. Intel-specific features: SGX, EPID, integrations with Intel platforms and software

Intel will have the lead on this work.  We expect that other vendor requirements will emerge as the community grows.

10. Python 3 readiness
11. Operating System independence

These are both long term goals that we can start working towards gradually.  

Python 2 support will be gone in a few years.

We have already had multiple requests for supporting other Host operating systems and we expect that to continue.  Moving in toward OS independence  means de-coupling the services from the OS which makes the services themselves easier to work on and more interesting for other projects and communities to use.  This will get easier the more OS patches we get upstream.




_______________________________________________
Starlingx-discuss mailing list
Starlingx-discuss at lists.starlingx.io
http://lists.starlingx.io/cgi-bin/mailman/listinfo/starlingx-discuss


More information about the Starlingx-discuss mailing list