Contributed by Adam Roberts, CDF Ambassador
Hey everyone, so we’ve just had our very first cdCon and I volunteered to co-moderate our Tekton Birds of a Feather session with Dan Lorenc (of Tekton fame) – and after it finished I honestly thought to myself... “that was quite possibly the most work-related fun I’ve had all year!”. And it didn’t even involve free food…
In this blog post, I’m going to explain why, and pass on a few acknowledgements and thank yous.
Attendance: over 800 active registrants out of 1300 signing up
Tell me more: https://events.linuxfoundation.org/cdcon/
You can find the full agenda and schedule above, but my take on things is that it was the first-ever cdCon, the agenda was filled with interesting experts in their fields talking about all things CI/CD, with useful sprinklings of ML and community-based talks, and the HopIn platform was used.
While I’ve only done one virtual conference before (spinnaker.live) I personally found using this relatively new medium of online conferences both interesting and productive.
I get to sit back and answer questions from the comfort of my home desk (more questions that I’d be able to manage in person, with folks queueing post-talk), and I can easily join sessions with a few clicks – or leave those that turn out aren’t a good fit for me, without any hassle or awkward “ahem”s, “excuse me”s, “thank you”s. The whole “this talk is full, sorry!” goes out of the window too.
From my perspective, the conference ran smoothly without issues – and I personally look forward to checking out the replays once they’re made available on the Continuous Delivery Foundation’s YouTube channel.
The BoF session (what is a BoF anyway?)
Wiktionary states: “Noun. birds-of-a-feather session (plural birds-of-a-feather sessions) (computing) An informal discussion group, typically at a conference, based on a shared interest and without any fixed agenda.”
Well, this is exactly what happened – and because we were all in using the HopIn platform from our own workspaces, we got to see some interesting cameos (I do believe I saw a standing desk, was complimented for my very much work-in-progress home gym set up, and we had a brief canine appearance, which is always a bonus)
Dan and I joined and, with myself always being the prepared talk, I was ready to pepper him with nine questions that I thought would be interesting for everyone to learn the answers to.
We made it through one of those questions.
Within minutes we had our guests arriving and joining us with their webcams and microphones on, so my other eight questions were well and truly tossed aside…
The folks we got to chat and engage with
What exactly was discussed?
In the talk I delivered on How the Tekton Community is Driving its Growth and Adoption, an audience participant named René asked about this, perhaps hinting at that’s a reason why Tekton isn’t so commonly adopted. I reached out to Vincent Demeester from the Tekton community straight away on Slack and he came up with the following doc, which we discussed at the next working group meeting.
Hint: you can find all of our Tekton working group meeting notes linked to from here.
Simon got really excited when Dan mentioned this as an idea for Tekton. You can check out https://landing.google.com/sre/sre-book/chapters/release-engineering/ for the background reading, and for Tekton itself check out https://github.com/tektoncd/community/pull/204. “A Hermetic Build is a release engineering best practice for increasing the reliability and consistency of software builds. They are self-contained and do not depend on anything outside of the build environment. This means they do not have network access, and cannot fetch dependencies at runtime.”
DSLs for Tekton Pipelines (or how about using GPT-3)
- Dan’s still on the waiting list for GPT-3, but we agreed nobody really likes to write YAML
- BlueOcean plugin was mentioned for writing Jenkins Pipelines
- Could Jenkins-X make this easier? Does it?
The challenges with Tekton
Adam mentioned sensible security constraints – todo get more details
Microservices that automatically intelligently update
IMO the most exciting of topics. I believe what we discussed was…
Microservices automatically receiving an update that their image to use has been updated, the newest image is then pulled in and used in a canary way, and if everything goes well, the change is persisted.
So there’d be a polling mechanism (a controller and this would all be operator/custom-resource driven?) somewhere for the image coordinates used by managed microservice deployments, this would go and update the Deployment Spec and set off some testing. I don’t know of something that does this already, but Simon brought up Razee and I thought about iter8.
It’s worth noting that both Tracy and Simon’s eyes lit-up while they were discussing this, so perhaps we’ll see something exciting in this space being surfaced.
My personal highlights
- The first time moderating anything. Pretty straight-forward…a few interesting questions in the back pocket, and trying to get everyone’s opinion heard and recognised
- Getting an insight into people’s living/work arrangements – including a participant’s comment about my very much work in progress home gym in the background (while himself using a standup desk). Thanks, Yardiel, you were a joy to chat with
- Gaining insights into what real end-users think of Tekton!
Ideas for improvements
- Would have loved to have even more participants pitching in – the platform itself, I’m sure, supports much more than we had. The awesome Jenkins X BoF had around 65 folks at once, and the chat was also super active. Having said that, there was never a dull moment with our session and we spoke consistently and interestingly for the entirety of the one hour allocated duration
- Would have been excellent if we had more diverse backgrounds joining – not just in terms of demographics but in terms of industry. Our Director of Community at the CDF, Jacqueline Salinas, has suggested this – if we could bring onboard folks from various different industries with their own ideas and perspectives we can all improve together
- Having the Jenkins-X folks present! Jenkins-X was mentioned a lot so having those who built the platform would have been immensely useful, availability permitting.
More positives and closing remarks
- Great ideas were definitely shared, hopefully to become more tangible in the future.
- People being honest and upfront. Challenges discussed in a productive manner, as opposed to solely having complaints with no solution or ideas presented by anyone) – made easier by having knowledgeable and appropriate participants
- No awkward moments. It would have been easy to have these and for all manners of reasons (technical difficulties, a lack of points to discuss, nobody having answers to questions…).
In closing, I would like to say a huge thanks to everyone involved in making this happen – from the speakers, the organisers, the attendees, and the folks in the speaker green room. I had great fun, learned a thing or two about a thing or two, and I’m hoping everyone else did too. Oh, and to Jacque and Roxanne for both the blog idea and the edits!
Now to check out those replays…