All Day DevOps 2018: Top 10 talks for dev and ops pros

Attending the All Day DevOps online conference is a bit like running a DevOps telethon. Over a 24-hour period, starting at 3 a.m. on October 17, more than 100 DevOps practitioners and IT leaders will take the stage at this free, annual online event. Last year, over 33,000 people attended.

This year's event, All Day DevOps 201, kicks off at 3 a.m. Eastern Standard Time on Oct. 17, with 125 back-to-back sessions. The content is widely respected: More than 120 companies have arranged group viewing parties during the program, which boast more than 250 supporter groups.

But with so many choices, how do you prioritize? TechBeacon has done the work for you. Here are 10 sessions you won't want to miss. Can't get it together for that 4 a.m. presentation? No worries. All sessions will be recorded and available for registered attendees to view immediately after each presentation ends.

How to Build a DevOps Toolchain That Scales

1. Keynote: Girl on a Mission: Stories and lessons from an out-of-this-world career in tech

In her keynote presentation, Cindy Healy, one of the 25 people who wrote the source code for the Mars Pathfinder, will share what she learned working on the project and discuss how to make a successful mission wherever you are. She'll discuss how she thrived on challenge and stepped up to do things even when she wasn't the expert.

"Cindy is going be talking about not being afraid to do something, to say yes, and then turn around and learn how to do it," said All Day DevOps co-founder Mark Miller, who is also senior storyteller and DevOps community advocate at Sonatype.

"My story is a feel-good story about technology and more. It is about some of the twists and turns, highs and lows that happen as you overcome the odds and achieve what was thought to be impossible," said Healy, director of Microsoft Worldwide Learning Experiences.

"As an award-winning storyteller, STEM advocate, and diversity champion, I share my story about my unusual career in technology through the lens of the life lessons learned as a member of the history-making Mars Pathfinder team."
Cindy Healy

2. Get started with DevOps in government

If you want to learn how to pave the way for DevOps in a process-oriented enterprise organization, then be sure to attend this presentation by Mieke Deenen, an independent consultant. She is helping the Dutch governmental organization responsible for social security benefits implement DevOps.

Deenen will offer insight into the organization's journey to DevOps, the many challenges the organization faced, and how it managed them. She will also offer some advice about how to use enterprise culture and policy to strengthen and embed your movement to DevOps.

"Mieke, who works with the Dutch Tax Authority, will talk about bringing DevOps practices into large government organizations," said All Day DevOps co-founder Derek Weeks, who is also vice president and DevOps advocate at Sonatype. "And she will also talk about the resistance that some organizations or people in the organizations feel, and how to maneuver through that resistance."

3. Keynote: Mapping value: The pros & cons of the manufacturing paradigm

The usual manufacturing metaphor applied to IT is the production line. This has served some elements of software delivery quite well. But the wider IT operating model involves many interconnected streams that don't fit into a linear model. The whole enterprise is the value chain, and that value-stream thinking must apply to the whole IT group and not just to software delivery.

In this keynote session, presenters Rob England, IT consultant at Two Hills Ltd., and Mark Whiting, founding partner of ValueFlow, will discuss the current thinking on value chains and value networks. They'll also talk about IT as a value creator (optimized for value or flow) versus IT as utility provider (optimized for cost or resource utilization).

"There is a tension between the 'factory' model and the need to remain agile (in the nimble sense), or non-monolithic. Automation increases agility to a point, then reduces agility, as Tesla found to their cost. We must understand IT value as a network and know when the value stream model, the factory paradigm, the theory of constraints, and even lean limit us instead of aiding us."
Rob England and Mark Whiting

4. DevOps at the Federal Reserve Bank of New York

In this session, Colin Wynd will discuss how the Federal Reserve Bank of New York (FRBNY) moves workloads to cloud environments while continuing to maintain the security levels needed for managing trillions of dollars. Wynd will offer an overview of the types of technologies for continuous integration/continuous delivery and how FRBNY is delivering solutions to its customers.

"This is a very heavily regulated environment, where they are moving trillions of dollars around between banks," Weeks said. "And they have to get that right in terms of the processes that are required to move and manage that money between the different banking organizations. But when they put this stuff up in the cloud, they also need to think about how to continue to meet regulations and security protocols within these environments as we move there."

5. The gas, the brake: Finding common ground with security folk

With a nod to Benjamin Franklin, information security expert Angela Gunn is of the mind that executives, developers, bug hunters, and compliance professionals must all hang together or hang separately.

Gunn is an incident-response consultant at BAE Systems Applied Intelligence. In her session, she will offer some perspective on how DevOps staffers, who likely feel that security experts would prefer cars with two brakes and no gas pedal, and security pros, who are convinced that DevOps is an excuse to joyride over their last good nerves, can find better ways to coordinate their efforts behind the wheel.

"I used to talk a lot about how security folk were different from other tech folk in that we had an essentially pessimistic view of the world. Instead of better-faster-cheaper, we live in a world where [security pros] are forever battling entities that never stop improving and never give up," Gunn said. "The truth, of course, is more complex." 

"Devs don't want to break the world and leave messes for someone else to clean up, and infosec folk don't want everyone to unplug all the things and move into caves. Safety and delight don't have to be at cross purposes, but to get to a place where those are harmonized, both sides need common ground and common language."
Angela Gunn

6. Developing for deterministic deliveries

The craft of producing quality software doesn't happen accidentally, according to Mykel Alvis, the DevOps coach for Cotiviti's R&D division. In his presentation, Alvis will explain how decades of engineering efforts, primarily in the manufacturing arena, have provided guidance for producing repeatable, reliable outcomes. He'll also discuss the importance of the concept of "immutable infrastructure" for DevOps.

"It's very specific to our process of software release, which uses the Maven release process as its format. The 'deterministic' element is around the use of an artifact repository to store dependencies for later consumption, as well as having an automated process by which those dependencies are consumed during the production of a release."

"My talk is about executing a repeatable process to produce software."
Mykel Alvis

7. Keynote: What is SRE, and why every DevOps professional should care

One of the hottest trends in operations today is SRE (site reliability engineering). If you want to know how to explain it to your boss, understand how it relates to DevOps, and know who can and cannot adopt it, then put this session on your calendar.

Attendees at this session will hear Dave Rensin, Google’s SRE director and the founder of customer reliability engineering (CRE), explain how SRE and CRE started at Google, why the company is working so hard to ensure everyone can adopt it, and why it should matter to every DevOps practitioner.

"The attendees are all familiar with DevOps, of course, and maybe some of them have heard something about SRE," Rensin said. "There is a misconception in the industry that somehow the two approaches compete."

"My talk will introduce SRE to those who haven't been exposed to it yet, explain how SRE and DevOps reinforce each other, and show how companies can start adopting SRE practices right now for super-high reliability."
Dave Rensin

8. Monitoring the easy way

Observability is a hot new buzzword that comes from control theory, which isn't so new or hot. In this presentation, Daniel Barker, chief architect at the National Association of Insurance Commissioners, will explain how to get more observability into your systems using Prometheus, Jaeger, OpenTracing, and Istio. Barker will also walk you through a demo and deployment using the operator pattern in Kubernetes.

"I will have 30 minutes to discuss a rather complex topic and do a demo, so we'll be moving fast. We're going to learn what observability is as it relates to control theory, what the different components of observability are within the bounds of technological systems, and how to deploy these to Kubernetes."
Daniel Barker

There will then be a demo to show a couple of failure scenarios that you'll be able to detect using Jaeger. "We'll also show how to deploy Prometheus in a way that allows simple management."

9. Mother Nature's Development Lifecycles, or Why The T-Rex Didn't Get Hand Extenders

We broke Mother Nature's SDLC (systems development lifecycle) for humans, according to Chris Roberts, chief of adversarial research and engineering at Lares Consulting. And we've been doing ad hoc rapid prototyping and flying by the seat of our pants for enough time that we've changed the course of evolution. In his session, Roberts will explain that unless we slow down a little and conduct proper analysis, design, and testing, we're going off the edge of the cliff at full speed.

"I think we'll have a little bit of fun in the session but also drive home some of the necessary messaging around collaboration, communication, and how we (those weird folks in security, IT, etc.) need to get our arses in gear and actually talk in a language that the business understands and what/how we do with these things called metrics. We'll also take a look at technology, throw most of it out, and focus on fixing the basics, which most folks ignore. Chances are we'll annoy a few vendors, lawyers, federal agents, and other folks along the way."
Chris Roberts

10. The Trust Algorithm For DevSecOps

There's a massive lack of trust between security and development at most organizations. However, a DevSecOps initiative is doomed without mutual trust. In this talk, Larry Maccherone, who leads DevSecOps transformation at Comcast, will introduce a set of steps, an "algorithm," to optimize the trust formula in the context of a DevSecOps initiative.

"Technology is easy. People are hard. You can have the very best technical solution, but if no one adopts it, then you’ve provided no value. My superpower—and everyone has one or more superpowers—is figuring out how to get software developers to change their behavior and truly adopt the things that us 'experts' are recommending. The people part of technology."
Larry Maccherone

Maccherone said his talk is for you if:

  • You're already struggling with getting development teams to adopt DevSecOps (or really any other) practices, or if you've been avoiding it and just focusing on the technical aspect.
  • Your security group says, "Those developers are just putting out crap that's going to get us hacked," and your development group says, "Security just gives us more work that prevents us from shipping value. It's easier to ignore and avoid them."
  • You're a student of human bias and how to overcome it, or you want to be one.

For the full list of sessions, see the agenda at the All Day DevOps 2018 online conference website. Registration is free. All presentations will be live-streamed, and available on YouTube following the conference.

Topics: DevOps