Yesterday concluded the 5th DevOps Days Vancouver. This is a yearly event that has been held since 2013 (excluding 2015). The format has remained consistent and inline with the DevOps Days conferences that have been occurring around the world since it began in Ghent, Belgium in 2009. Hot of this heals of this event are others around the world...
- Apr 24 - 25: Seattle
- Apr 24 - 25: Copenhagen
- Apr 24 - 25: Tokyo
- Apr 26 - 27: Jakarta
My personal relationship with DevOps Days is a little different to most, having been the lead organizer of the first of these conferences held in Vancouver and Canada. Having unintentionally got myself into that position and realizing how much work it was to run a conference, I was happy to pass the reigns the following year. I definitely appreciate all the efforts that Kelly Shkuratoff and the rest of the organizers put into these conferences. The conference this year was great!
For those unfamiliar, the conference is a two day event and is split between talks in the first half of each day and open-spaces sessions during the second half of each day.
I won't list all the talks, but some that resonated with me were...
Heidi Waterhouse: Y2K and Other Disappointing Disasters
Heidi opened the conference with a look at disasters that didn't happen and the reasons they didn't happen. Heidi has had firsthand experience of the effort that went into preventing her organization from suffering a Y2K outage. She also brought an awesome collection of stickers from the many conferences she attends in her role as Developer Advocate at Launch Darkly.
Amin Yazdani: Psychology of Estimation: Faster is not always better
Based on the book Thinking Fast And Slow by Daniel Kahneman, Amin summarized some of the bad project planning we do when our brains assume they know more than they do and plan for best case scenarios. I've heard of this book a few times, but Amin's talk convinced me to order it and thanks to Amazon Prime it's in my hands already and partially consumed!
Kyle Young: On the Virtues of Being T-Shaped
Kyle's lightning talk suggested that we can benefit more when our shape as engineers leans more towards breadth than depth. He also believes that it's easier to attain depth in one area when you already have experiences in many other areas.
Jeremy Pierre: Developer to Manager (and back, and back again)
Having gone back and forth between pure engineer and management roles, Jeremy's story had many parallels to mine. As he now looks at a management role, his self-reflection and research led to some good insights that I can relate to and a good list of books to read.
- The Manager's Path by Camille Fournier (I've read and recommend)
- Ego Is the Enemy by Ryan Holiday
- Nonviolent Communication by Marshall B. Rosenberg PhD
- High Output Management by Andrew S. Grove
- Managing Oneself by Peter F. Drucker
Will Whittaker: Bitrot: A Story of Maintenance Failure
Will is one of the founders of the first company I worked for in Vancouver, MailChannels. He's also one of the organizers of the security conference CanSecWest, for which he created a registration and payments system. He detailed the long history of this overly ambitious and badly maintained project. This was reminiscent of technologies I also used over decade ago, which I have since forgotten, such as specific Perl MVC frameworks. It made me glad I'm not still supporting them now.
Sean Porter: The Art Of Monitoring
This was great succinct history of monitoring tooling in lightning talk format.
John Rauser: Dealing with Dark Debt in your Culture
Culture is a key ingredient of these conferences, since empathy is one of the underlying tenets which enables Dev and Ops to work as one. John framed our inability to quantify cultural rot in the same way we view technical debt and provided some frameworks for managing it. An interesting point was that debt should not be completely avoided, since debt it generally used to make things go faster (e.g. VC investment, technical spikes of work). You do need to understand where cultural debt is creeping in, plan for it and manage it.
Open-spaces I attended were related to...
I have no experience Kafka, but always interested in hearing what people are doing with this technology. Segment and Salesforce brought a lot to the discussion and some impressive numbers, too.
GraphQL and API Gateways
Something we're investigating at Bench is GraphQL. It was good to learn how others were using this and some of the benefits they were getting. Centralized authentication was one benefit I'd not thought of before and I have definitely seen the odd REST endpoint that forget to add auth. GraphQL on first glance, seems like something that will cause more headaches than it's worth, but that wasn't the impression I got from the folks actually using it.
Having started down the road of Kubernetes, it was good to hear from others at different points in the journey. Definitely a lot more words of caution than I was expecting here, but we outlined a good set of safeguards during this discussion.
Key and Secret Storage
Having invested in rolling out Hashicorp Vault and other technologies, it was good to get confirmation that this was the right direction and share our experiences here.
...and that was DevOps Days 2018
A video stream of all the talks is available online at https://www.devopsdays.org/events/2018-vancouver/livestream/
I'm looking forward to DevOps Days Vancouver 2019!