PTG Denver Survey Responses
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

View only
What is the most important thing we should do to improve the OpenStack software over the next year?What is one change that would make you happier as an OpenStack Upstream developer? 
Open-Ended ResponseOpen-Ended Response
Continue to push connecting developers with operators.Less distractions from downstream work.
Stability, stability and stability. (With the community shrinking, one of the side effects seems to be gaining more of that as there is less orgs/people insisting their cornercase new feature being on the next release.)Getting rid of these "One size fits all" artificial community goals. We had fairly tight cycle so the community goal discussion kind of slipped from us. The approach and the reasoning to move the policies to the code is pretty horrific, causes security concerns and we can't do it consistently anyways. But I guess someone had to be able to justify their position in the employing organization.
Keep focusing on good documentation.??
get more developer.translator.
Loaded question! There will always be lots of things worth doing. But I suppose one pain point folks are having is in attracting folks with sufficient talent/commitment. It's easy for companies in particular to not regard open source contribution/review capability as a full-time job. Perhaps more educational/evangelical outreach around all of this?If we keep doing that singalong at the PTG, I think we'll all be just fine. :)
Ease of skip upgradeMore structured information on what is going on/needs contribution
Go back to focusing on the developers as rockstars that actually get the stuff done that will keep openstack moving. Stop choosing stupid sessions for the Summit and focus on what is actually being done.Stop treating us like free labor to execute the things that the Telcos and Users want without context or funding for actual work.
Continue to iterate as has been the way for awhile now. Incremental improvements are the best way forwardMore reviews!? :-) The community is great and responsive, but it would be great to have more involvement.
Improve work / feedback loops with other communities More love for the infra team and all they do :)
Fixing bugs and removing sharp edges.Simpler setup of CI infrastructure (this might be completely impossible)
More stabilization, lowering the entry barriers -> attracting more clients, keeping up the hype -> will make the 7) more easy for me ;-) Like the simplification website mockups by Mike Perez!Getting more time for upstream work from my very own company ;-)
Try to prevent laying off the coresI don't know
I don't know. I think you're doing great. I'm pretty happy.
Prioritize HardeningDunno
Groom more Dev and reviewers.
Actually policy in code is a pretty awesome goal already.Don't glorify the TC/Infra/Release people.
Ask me next time - just getting acclimated to the Openstack Cinder teamI plan to help with documentation. Again I'll have a more informed opinion in March 18
Keep it going. :)Maybe extend the schedule to allow for better cross team meeting.
Try to make bug list looks smallPTG, it has been very helpful for discussing our plans, finalizing it and if something went wrong then we can revisit the same during summit.
ContainersMiss the summits! :)
improve upgrades testingmake it working outside of devstack :-)
Get the basics right. Improve performance.This is my first PTG and I hope my company will continue funding my attending future events.
Be responsive to bugs and operator feedbackI preferred the old model of the design summit and letting projects have a midcycle on their own. It was easier and less expensive for me
Improve and expand the user documentation.Having reviewers relax a bit and submit follow-up patches for minor issues with patches from more casual contributors. Not to lower standards, but to allow people who aren't 100% upstream to be more effective.
Stay positiveBrief reports from project teams
Stabilise.Nothing comes to mind.
Securing more cloud resources for the CI : when its slow, this slows down our workflow.Zuul v3. A gerrit bot that would apply basic heuristics to propose people to add as reviewers of a change, e.g. based on the people that have touched the same files in the past.
Increase scalabilityOverall community gathering
Focus more on stability and less on features.More complete CI testing to reduce regressions.
No responseTwo comments about PTG&Summit co-location: 1. Back-to-back events in the same city may not be most people's first choice, but it *might* be most people's second choice and a good compromise between the old co-location and the new PTG. I don't remember if the surveys included ranked preferences, but if not, it may be useful. For people who attend only one of two events, there would be little downside, with only the upside of greater opportunity for the two audiences to interact at the overlap events. For the people who attend both (myself included), it may be a grueling and undesirable situation, but we are the minority, and at least we do save ourselves from an extra plane trip. 2. In terms of cost, it may be possible to do the PTG portion at a cheaper suburb of the summit city. It'll definitely add burden on the organizer, but it's a thought.
continue forward momentum More piano bar karaoke ;)
Continuing to improve the operator experience through better deployment orchestration.N/A
A sense of discipline could be brought to how technology pieces of OpenStack evolve, overall.More, methodical training. Perhaps an online crash-course. I could then easily bring more of my team along into the upstream developer fold.
Make it simpler or at least more intuitive to deploy and use. With great power comes great complexity and we should ideally keep the power but simplify the bootstrapping and reduce the operational costs..
put it right after summit eventput ptg right after summit
make it friendly to the usersfreedom
On boarding more contributors and coresPTG at a fun place
ease of use, container integrationCI setup
Improve ops toolingBan eventlet :-)
Stability and performanceUnsure
Containers make it prioirtyMore recognition
Keep key teams that enable development and that are losing participation going (infra/interop/refstack/QA). No idea.
Be open to new ideas.LOCI!
More stability, better testing, better QAIt takes to dang long to report/fix simple bugs. that's more of an issue with the OSIC disruption but it's still a major pain point.
I'm still too new to have a strong opinionSame -- still new. Haven't quite figured out what's growing pains for me and what are long term issues that I want to complain about :-)
Better guidelines for updating... its very common to talk to clients using three cycles behind versionsNothing special
Evolution in baremetal support.Not upstream developer.
OpenStack's ContainerizationHaving more operators feedback
Get folks to focus on ease of deployment and integration with stuff outside of openstack.more defined and centralized agendas for the projects. perhaps publishing them centrally instead of having separate etherpads.
Fix RBACexpanding gerrit review comments causes the browser (at least Firefox and Chrome on Mac) to jump elsewhere on the page and then you have to scroll back to see it. Very annoying, especially if there are a lot of comments.
better team organizationtimely review
continue to focus on maturity-related things like upgrades, docs, stability, gating.addition of more realistic reference deployments in gates.
Get people to continue community wide strategic contributions, rather than focus only on company specific small contributions.Getting friends who left the community back into the community.
Improving our interaction and interoperability with container solutions.Being able to make it to more of the OpenStack Days events/help with more of the Upstream Institute training. That isn't really something you guys can fix though. ;-)
Make it easily consumable. Stop adding new services now and drop useless services that have no reason to be in openstack.More contributors.
Quality & robustness not sure
Stablization. Easier upgradesMore dev hardware
eliminating complexityStop relying on devstack in the gate for most projects
StabilizationMore contributors
network accessibility. hotel network accesibilitypower supply
ease of use. more clear integration points between projects.
Streamline projects to have one solution per technology.More events like PTG which gives developers more face to face time.
I think Openstack should focus on stability and performance.I can discuss and contribute to Openstack software with engineer over the word.
OpenStack should elaborate container, BM, VM interoperability.A little improvement is gate infrastructure is more stable and easier to debug.
deprioritize the use of devstack for more realistic customer like deploymentsseeing more of the projects check jobs with a wider variety of deployment tooks like openstack-ansible, tripleo etc.
Continue hardening the software and make sure there is plenty of consideration in the container spaceI am pretty happy
Cleaning up dead projects, polishing up documentation, making sure "major" projects have enough cores.A better feedback mechanism for project developers.
Consistency across projectsConsistency
Get the community closer together. SIGs and WGs need to increase and require a liason from each project. A lot of different projects are doing the same things differently and becoming inconsistent.Go back with Design Summits in the Summit and Midcycles. There will be less money spent on Travel Support for key people that have to travel 4 times a year.
Less politics.Less politics.
Network connectivity :DN/A
More focus on the end-user experience; encouraging companies sponsoring contributors to focus more on polishing the user experience and on bug fixing, and less on new featuresLess travel to the US
make sure projects stay healthy...
Have sessions shared amongst all upstream devs to review feature proposals at the start of each cycle (to get interested people involved into it)More/Faster CI infra
Have a better interproject direction.Perhaps more cores for the projects, as they are overloaded.
Clarify the packaging situation.Foundation support for remote conferencing.
DebugabilityMove CI to a more stable foundation (rather than devstack and grenade)
I think all goes mh good but main challenge will be to keep the same pace in OpenStack Development. Really need to thing about keeping more contributors in OpenStack.Keeping and maintaining the same pace and same motivation.
Simplify it :)I'm happy!
Maintain stable group of developersMaybe condense to 4 days, as days 1 and 2 were light
Unable to keep track of all the new projects that are spinning off at a rapid rate. Not sure I can offer a solution here yet, but how can this mushrooming be more organized or tamed?!Maybe each group can start off their sessions with a level set with a short summary of what was accomplished during the previous release.
Continue to stabilize and reduce complexity.More inclusiveness. Some projects have cliques that are hard to break into.
Figure out how to fit better into the evolving infrastructure choicesGet to a point where we have a stable set of people, expectations and plans.
Simplify and tighter integration with K8sSlack IRC gateway for #openstack- channels
Sharper focus on the fundamental aspects of cloud operation.Moving to Zuul v3!
Stability, deployablity, and health monitoring to ensure that your cloud is running and performing well. All projects include a logger which emits performance metrics
Location isn't great. Hotel renovations and train are distracting. The overall meeting has been good. Just a bad venue.
It's more around staying relevant. Everyone's so excited about Kubernetes, but they're often forgetting that deploying it without some kind of IaaS is a really frustrating experience.So far so good! Infrastructure works well and people are helpful.
Upgrades! We can't possibly handle skip upgrades until we can do lockstep upgradesFewer developers replying to with "not my project not my problem". We're all *OpenStack* developers.
We need companies to send more developers to join PTG(it's hard I know:))better equipment for video conferences:)
We need to keep all the different vendors we have, and try to get more, as losing a company means losing core reviewers.Having jenkins jobs be able to finish quicker.
Have *one* strategic upstream supported deployment tool which can handle upgrades. There are many different possibilities AFAIK, which creates uncertainties to where to bet on. The missing tooling is IMO the major inhibitor to adopting OpenStack at all and keeping up with the releases once you deployed it (see the amount of releases which are still in production but have reached upstream EOL).I can only talk about Nova here. The bottleneck there is the amount of reviewers capable of reviewing and merging the changes. I sometimes have to wait for weeks until a review gets done. I have no solution to that bottleneck TBH, but it would make me happier if we could find a solution for that. This topic usually comes up at every Summit/PTG and there is no solution at the horizon.
Greater support for componentised deployments (subsets of services for non-compute-cloud needs).Weaker identity for specific project teams: more cross-team code contribution.
Main menu