I wanted to share my responses to a set of questions from the Product
Working Group who is working with the OpenStack Foundation to collect
updated information on the Newton release for each project (along with
directional guidance on future release plans).
This is a follow up at the end of the release based on an interview I
gave toward the beginning (shortly after the Austin summit):
https://www.youtube.com/watch?v=EHCj1DZNNWw
The first one is hard to answer based on our list of priorities:
https://specs.openstack.org/openstack/nova-specs/priorities/newton-priorities.html
The immediate user POV features from that list in Newton that we
completed were API policy defaults in code, glance v2 integration and
Get Me a Network. However, more work went into cells v2 and scheduler
(placement API) changes, but the cells v2 and placement API features are
still optional in Newton, and are really operational features. So it's a
hard one to answer but I just left what was in the initial interview
responses.
For the Themes section I gave a lot of info, maybe more than they wanted
for marketing slides, but the Nova team got quite a bit done in Newton
and I'm not including everything, but it'd be remiss of me to not
include a lot of these.
As for Ocata and Pike, we don't determine the priorities for those until
the summit, so they are just educated guesses at this point.
Note: these are already submitted. I'm posting here for informational
purposes to let people know that sausage is being made.
Questions:
1. What are the top 3 or 4 features that your team is delivering as part
of the Newton release?
- Limited cells v2 support
- Placement API (tracking quantitative resources)
- API policy defaults in code
- api-ref docs cleanup
2. Areas of focus/ themes you will target (please give an example of how
your team will execute on these themes in the Newton release)
Theme 1: Scalability [Changes that will impact the scale at which the
service can operate]
- Cells v2
- Placement API
Theme 2: Resiliency [Changes that will improve the availability of or
ability to recover from failures for the service]
- Ironic multiple compute host support for HA
Theme 3: Manageability [Changes that promote operational ease-of-use or
benefits operators in general]
- API policy defaults in code
- Added support for libvirt live migration post-copy mode
- New vendor data API:
http://docs.openstack.org/develope//nova/vendordata.html#deployer-provided-data
- Mutable debug logging configuration.
- nova-manage command to synchronize quota usage
- Improved configuration option documentation
- A subset of notifications are now versioned
Theme 4: Modularity [Changes that help the service architecture divide
into more manageable/smaller services/code-base]
- os-vif library integration for LinuxBridge and OVS on libvirt
computes
- Castellan library integration for key management
Theme 5: Interoperability [Changes that help enable the service to
operate across multiple OpenStack clouds [federation] or promote a
common experience when the service is deployed in separate
OpenStack-Powered clouds [interop]]
- Proxy resource APIs are deprecated in the v2.36 microversion:
http://docs.openstack.org/developer/nova/api_microversion_history.html#id33
- Glance v2 integration
- Removed support for API extensions
Theme 6: User Experience [Changes that benefit end-users of your service
-- not the operators but the people consuming the service]
- api-ref docs cleanup: http://developer.openstack.org/api-ref/compute/
- Server tag support - v2.26 microversion:
http://docs.openstack.org/developer/nova/api_microversion_history.html#id23
- Virtual device tagging - v2.32 microversion:
http://docs.openstack.org/developer/nova/api_microversion_history.html#id29
- Get Me a Network (auto-allocated networking topology) - v2.37
microversion:
http://docs.openstack.org/developer/nova/api_microversion_history.html#id34
Theme 7: Security [Changes that improve security ]
- oslo.privsep integration for os-vif and os-brick libraries
- Multi-tenant networking support for Ironic computes
3. What are the top 3 new features planned for development in the [Ocata
release]?
New features/ enhancements
- Multi-v2 cell deployment support
- Placement API used for scheduling decisions
- libvirt storage pools
- API discoverability - What can this cloud do? What am I allowed
to do on this cloud?
Areas of focus/ themes you will target
- Scalability, manageability, interoperability and security
4. What do you project will be areas of focus in [Pike release]? What
does your crystal ball show you?
New features/ enhancements
- Model resource provider capabilities and use them for scheduling
- Improved API workflows between Nova/Cinder and Nova/Neutron
Areas of focus/ themes you will target
- Scalability, manageability, modularity, user experience
5. What are the questions you would like to ask users for additional
insights to help guide the planned development in the Ocata release?
- Are API microversions working for you?
- Is it more important for Nova to fix bugs or add features?
- What is missing from the compute API?
- What is the most difficult thing about deploying/operating Nova?
- Is Nova effectively communicating changes?
--
Thanks,
Matt Riedemann
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
