The Hive at Monitorama Baltimore 2019


As a human who entered this space via marketing, I often turn to my teammates for context. I recently asked the hive, why do we like Monitorama so much? Briefly:

  • well curated community-driven event
  • good mix of new and experienced speakers
  • organizers really research trends of the industry
  • excellent hallway track that is easier to manage (as an attendee) because event is single track

I got a nice fountain of observations from the team via Slack. Those were insightful, but now that I’ve actually been to (a) Monitorama, I get it. This event feels like a reunion. Folks come together on common ground, but from many different experiences. I’d wager that many folks who came to their first Monitorama in Baltimore this past October left feeling more connected and included in the community.

The talks

Which leads me to Pete Cheslock [currently unaffiliated]. In Untitled Monitoring Talk, Pete covers how monitoring has evolved just in the few years since Monitorama began. Folks come in at all different levels and from different backgrounds, and there is just so much context to the monitoring space, it can be difficult to navigate.

Pete’s roll of other potential titles helped me understand more:

  • have we finally reached #monitoringlove; but if we have why does logging still suck?
  • all logs are wrong, but some are useful
  • monitoring: the bad parts
  • I just took a dna test; turns out I’m 100% sick of logs
One of Pete’s slides 🙂

Pete’s talk is one reason I feel that sense of inclusion. His story helped me feel the pain that can come from trying to make sense of complex systems. ‘The future is here; it’s just unevenly distributed.’

If you’re like me, and you could use a big dose of context and connection to the emotional pain of logging, do not miss this overview. Thank you, Pete, from the bottom of my gin and tonic.

With a journey from no observability to practicing it at gigantic scale, Philip O’Toole’s [Engineering Manager | Google] talk Observing Observability covers why observability systems struggle and why they are hard to build. He explains the challenges tools have keeping up with problems, ponders why so many companies are building observability tools, and shares lessons for how to think about tools and teams. His lessons, quickly:

  1. logs and time series are different
  2. everyone thinks a new query language is the solution
  3. the nature of innovation
  4. it’s a crowded market
  5. the product development challenge

"It may be that too many of us are building what we know..."

My favorite quotes in this talk explain why so many ‘observability’ products may be burgeoning: “Engineers want to solve the problems they understand best,“ and, “It may be that too many of us are building what we know… maybe some of us should look outside solving our own problems because there’s so much more of the world that needs to have software applied to it.“ I appreciate the experience Philip brings to light in this presentation.

Coalescing systems data without losing fidelity is a thoughtful breakdown of how teams can get what they need from their data without drowning in the floods of collection and cost. In tandem with the artistically expressive @emilywithcurls, Liz Fong-Jones [Developer Advocate | Honeycomb] brings a intelligent, technical approach for reducing the cost of storing data to an accessible level.

Liz Fong-Jones speaking

Liz’s talk directs the focus for software development toward user happiness. Aimed at that goal, she outlines data-driven approaches for managing data to answer questions beyond break-fix and to understand system behavior. SLOs are front-and-center: how to think about them, measure them, and build them. Recognizing that full context of data (as opposed to aggregates of aggregates) lets dev teams know exactly how production is working. Understanding the kinds of questions you can ask when you can easily navigate raw data moves teams beyond the patterns they already know.

The hallway track

An audience member stopped by the Honeycomb booth to share his key takeaway from Liz’s presentation, “Don’t lose your marbles; don’t aggregate.” We whole-heartedly agree: context is everything.

Applause to Jason Dixon and the whole organizing team for a great event. I appreciate that Monitorama BWI chose Baltimore Center Stage as its venue, meaning monies went to the arts, and brought in local vendors for snacks and great coffee. Thanks for continuing a community event that indeed gives to the surrounding community. And thanks for making me feel welcome, too.

Want to join the growing observability community? Sign up for a free Honeycomb trial and you’ll be invited to our private Slack to learn and share with the very best in the industry!