I sent a similar email after Pike was released [1] and these are our blueprint burndown chart results for Queens [2].

Comparing to Pike, the trends are similar, with the overall numbers down. Things ramp up until the spec freeze, then tail off, with a little spike toward the end to get things in by feature freeze.

Comparing final numbers to Pike

Max targeted / approved for Pike: 76 / 69

Max targeted / approved for Queens: 65 / 53

Final completed for Pike: 50

Final completed for Queens: 42


Within targeted/approved/completed, the relative numbers are about the same and our percentage of approved / completed is actually better in Queens (72% completion percentage of approved blueprints in Pike compared to 79% completion percentage of approved blueprints in Queens). So while we targeted fewer blueprints in Queens, we did a better job of actually completing them. To me, this indicates some level of stability with the things we've been working on over the last several releases, particularly with respect to cells v2 and placement. It likely also means there are just fewer people / organizations trying to contribute, which isn't a surprise to me with the maturity of OpenStack. This can be both a good thing a bad thing, but I'm not going to try and get into that here.

We can talk about this and more during the PTG when we go through our Queens retrospective [3].

[1] http://lists.openstack.org/pipermail/openstack-dev/2017-September/121875.html [2] https://docs.google.com/spreadsheets/d/e/2PACX-1vRh5glbJ44-Ru2iARidNRa7uFfn2yjiRPjHIEQOc3Fjp5YDAlcMmXkYAEFW0WNhALl010T4rzyChuO9/pubhtml?gid=128173249&single=true
[3] https://etherpad.openstack.org/p/nova-queens-retrospective




OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe

Reply via email to