Summaries
POST /api-sig/news [0]
Release countdown for week R-14, November 18-24 [1]
[0] – http://lists.openstack.org/pipermail/openstack-dev/2017-November/124633.html
[1] – http://lists.openstack.org/pipermail/openstack-dev/2017-November/124631.html
Upstream Long Term Support Releases
The Sydney Summit had a very well attended and productive session [0] about to go about keeping a selection of past releases available and maintained for long term support (LTS).
In the past the community has asked for people who are interested in old branches being maintained for a long time to join the Stable Maintenance team with the premise that if the stable team grew, it could support more branches for longer periods. This has been repeated for about years and is not working.
This discussion is about allowing collaboration on patches beyond end of life (EOL) and enable whoever steps up to maintain longer lived branches to come up with a set of tests that actually match their needs with tests that would be less likely to bitrot due to changing OS/PYPI substrate. We need to lower expectations of what we’re likely to produce will get more brittle the older the branch gets. Any burden created by taking on more work is absorbed by people doing the work, as does not unduly impact the folks not interested in doing the work.
The idea is to continue the current stable policy more or less as it is. Development teams take responsibility of a couple of stable branches. At some point what we now call an EOL branch, instead of deleting it we would leave it open and establish a new team of people who want to continue to main that branch. It’s anticipated the members of those new teams are coming mostly from users and distributors. Not all branches are going to attract teams to maintain them, and that’s OK.
We will stop tagging these branches so the level of support they provide are understood. Backports and other fixes can be shared, but to consume them, a user will have to build their own packages.
Test jobs will run as they are, and the team that maintain the branch could decide how to deal with them. Fixing the jobs upstream where possible is preferred, but depending on who is maintaining the branch, the level of support they are actually providing and the nature of the breakage, removing individual tests or whole jobs is another option. Using third-party testing came up but is not required.
Policies for the new team being formed to maintain these older branches is being discussed in an etherpad [2].
Some feedback in the room expressed they to start one release a year who’s branch isn’t deleted after a year. Do one release a year and still keep N-2 stable releases around. We still do backports to all open stable branches. Basically do what we’re doing now, but once a year.
Discussion on this suggestion extended to the OpenStack SIG mailing list [1] suggesting that skip-release upgrades are a much better way to deal with upgrade pain than extending cycles. Releasing every year instead of a 6 months means our releases will contain more changes, and the upgrade could become more painful. We should be release often as we can and makes the upgrades less painful so versions can be skipped.
We have so far been able to find people to maintain stable branches for 12-18 months. Keep N-2 branches for annual releases open would mean extending that support period to 2+ years. If we’re going to do that, we need to address how we are going to retain contributors.
When you don’t release often enough, the pressure to get a patch “in” increases. Missing the boat and waiting for another year is not bearable.
[0] – https://etherpad.openstack.org/p/SYD-forum-upstream-lts-releases
[1] – http://lists.openstack.org/pipermail/openstack-sigs/2017-November/000149.html
[2] – https://etherpad.openstack.org/p/LTS-proposal
Quelle: openstack.org
Published by