Functional Specialty Teams - No longer active but keeping for reference
This matrix was drafted by Sundar Krishnamoorthy (Deactivated)and Jamey McCabe per 3/9 Roadmap meeting action item to start identifying topics we can collaborate on. It is intending to capture projects (and/or "Stories" now called "Development Proposals") that 3 or more members want to contribute to. X means that member is active or stated in the past they are ready to contribute. The Dark brown is more an 3 members and light brown 3 members. Red are items that came up in the past but may need to be removed due to lack of contribution.
| AT&T | Intel | SKT | R Jio | Comcast | NTT | Orange | Walmart | NOTES | |
|---|---|---|---|---|---|---|---|---|---|
| Containers | x | x | x | ||||||
Skip-level upgrading in place Backward/Forward Release Compatibility (N+1, N+2) | x | x | Seeks to jump ahead from an older release (Juno, Kilo) to a current release (Ocata...) in a hitless in place upgrade. How it can be done, etc. Seeks to assure that all releases will always be backward/forward compatible to ease upgrading in the scenario where large operator would upgrade 1x per year (each time jumping one release). | ||||||
| Kolla | ? | x | x | x | Doodle pad setup to explore availability for week of April 24-28 at: https://doodle.com/poll/pzu26y4ttqxhtaew Meeting page: Containerized Control Plane via Kolla - Intro | ||||
| OpenStack-Helm | x | x | x | ? | Schedule an OpenStack-Helm Overview session (dedicated meeting) Next Steps:
| ||||
| Salt Stack | x | ||||||||
| Dockerized Ceph | |||||||||
| (Generic) Containerization | x | x | x | x | x | x | What differentiates this item? Is this about containerizing workloads? Should we request a Forum session for this? Moderator? | ||
| Platform-wide Traceability | |||||||||
| Platform-wide Logging Error Codes | ? | ||||||||
| Operations and LMA | |||||||||
| SNAP | x | x | x | ||||||
| Ceilometer | |||||||||
| x | |||||||||
| Support for 1000 compute scale per openstack instance | x | Seeks to enable OpenStack to support 1000+ node scale AT&T - Kandan Kathirvel request | |||||||
| Networking | |||||||||
| Gluon | x | ||||||||
| BGP-VPN | x | ||||||||
| Network Function Chaining | ? | x | |||||||
| OpenContrail | x | x | x | ||||||
| Resource Management | |||||||||
| Blazar Project | x | ||||||||
| RBAC Moon | x | x | |||||||
| Openstack Resource Management (ORM) | x | x | x | How is this being interpreted? What is the specific focus here? | |||||
| Security | |||||||||
| Enterprise Security Policy Compliance LCOO-7 | x | Andy - Focus would be on defining the requirements necessary to comply with common enterprise security policies and pass security audits. This is a significant pain point for AT&T currently. OpenStack falls far short of meeting compliance requirements. There is an already existing Development Proposal (aka "Story) that has been abandoned which we could take ownership of. | |||||||
| Test Automation/CICD | |||||||||
| Role Based Access Control (RBAC) | x | x | |||||||
| Testing Cinder API | x | ||||||||
| Destructive Scale Testing (500+ nodes) | x | x | Refer to LCOO-4 LCOO-4 - Getting issue details... STATUS Sundar - 500 node testing in progress. test plan out. Can accept edits | ||||||
| Scale testing (1000+ nodes) | x | AT&T - Kandan Kathirvel request | |||||||
| Openstack destructive/extreme Testing LCOO-4 | x | x | x | x | x | AT&T - Interested. Draft a story and start from there. Asked for usage of "Chaos Monkey" Shintaro - compute node/control node failure and check if the controller/API will work under failure scenario (create a framework to do the testing in OpenStack infra or 3rd party CI) Pike timeframe. (To bring his team to discuss how its done) Jaesuk - Based on containerization. HA and failure test automated. Can collaborate with shintaro if he provides the framework. Aditya - Not the SME, agree with everything being said. CICD interest. In Agreement and in progress internally. He needs to connect with his team. Abdelhadi - Thinking of something that we are looking to do (rabbit MQ testing that are geo separated from control node 100-150ms delay) . Automate the tests and inject delays and monitor results. Megan - Get back to link in with the team. Involvement TBD. test plan could be a bit different based on what is tested - containerized openstack vs vanilla openstack Next Steps:
(start offline and then start a working session)
| |||
| Masakari Project (HA VM's) | x | ||||||||
| https://etherpad.openstack.org/p/LCOO-Working-161117 | |||||||||