Devopsdays Austin 2019

Practical steps for getting started with Chaos Engineering. Using concrete examples we'll cover onboarding teams onto a Chaos Engineering platform, identifying teams that are ready to do GameDays and creating feedback loops to measure resiliency.

Chaos Engineering is quickly gaining adoption in a number of companies, but there are still very real challenges to adoption. In this talk, we will discuss practical steps for getting your teams started with Chaos Engineering and rolling the practice out throughout your organization. Using concrete examples at Uber, Gremlin and Under Armour, we'll cover onboarding teams onto a Chaos Engineering platform, identifying teams that are ready to do GameDays and creating feedback loops so that improvements in resiliency can be measured and communicated across your organization.

About 6 years ago I had the pleasure of coming to Austin to speak at a conference for the first time. This talk was about how to retain your hardest to hire folks. It turns out 6 years is a lifetime in technology, let's revisit this talk, see what still works, and what can be improved.

Back in the spring of 2013, I had this idea floating around in my head. I had watched so many companies spend countless time and energy in order to recruit the best staff possible. But after you were in the door, it seems like none of them really cared about how you might retain all those people you spent years investing in. I had to pleasure of giving my first conference talk ever at DevOpsDays Austin about this topic of "how to retain your hardest to hire positions". Why do companies spend FAR more resources to hire new people, when losing any one of their existing employees would in many cases be a massive hit to the productivity of the company.

Now - 6 years has gone by. A lifetime in the technology space. To give you perspective - I gave this original talk BEFORE Docker was announced by Solomon Hykes at Pycon 2013. It can be hard to imagine a world before tools like Docker, Kubernetes, or even concepts like Serverless took over the day to day conversation. So much of HOW we get work done has changed in the last 6 years, has anything changed with how we recruit and retain our valuable employees?

Let's revisit this talk and see what still works, what needs improvement, and new learnings from the last few years.

Can Agile work for security teams? Let's discuss practical ways in which security teams can adopt agile processes and use DevOps tools to automate and track SOC tasks, as well as some of the challenges that can be encountered when attempting this endeavor. It's time for security to become Agile!

Security Operation Center (SOC) teams are responsible for protecting company assets by auditing, detecting, and remediating security threats in their networks and applications. As teams of developers evolve to put people over processes, and as DevOps makes its way into the core processes of many organizations, it becomes necessary for SOC teams to grow in that same direction to enable effective collaboration between security and development teams. In this talk, we will discuss our experiences using Agile and DevOps for building a SOC team. We will also consider how SOC teams can implement agile processes and use Azure DevOps to automate and track of SOC tasks, use Git for keeping track of security configurations, and document processes that enable more effective collaboration with DevOps teams in a practical way. Lastly, we will discuss some challenges that we have encountered when implementing Agile for SOC small teams in organizations that still rely on waterfall methodologies for most of their processes.

Do you know what the value in the marketplace is for your current skill set? Do you know if you are one of the many who can claim to have a pay gap? Attendees will learn not only how much money they should be making according to their skill set, but also what to do to close any existing pay gap.

Attendees will learn not only how much money they should be making in accordance with their skill set, but also what to do to close any pay gaps if one exists. There will also be some salary negotiation tips as well as tips on how to increase your market value.

This is not a problem that women and underrepresented folks are powerless over. We don't have to wait on politicians or companies to solve this problem for us. These are issues that we can solve for ourselves. Find out how at this talk.

Chaos Engineering is a term that has recently begun to be tossed around in Engineering circles with great frequency. This talk provides a high level overview of what "chaos engineering" is as a discipline, followed by a discussion of methodology used, and how to get to running your first gameday!

Chaos Engineering is a term that has recently begun to be tossed around in Engineering circles with great frequency. This talk provides a high level overview of what "chaos engineering" is as a discipline, followed by a discussion of methodology used, and how to get to running your first gameday! Other topics include: know what steps you can take to start doing CE "in the wild", realize the valuable outcomes of having a CE group at your org, prepare for common CE myths, and getting some resources for further investigation of the discipline.

Do you marvel at the idea of production-grade Deep Learning that can scale infinitely in nearly constant time?

Using a general purpose NLP application as an example, you will learn how to design, build and deploy deep learning applications using serverless computing.PS: BEWARE: Talk contains code!

This talk will discuss:

- The architecture and design of a serverless Deep Learning based system for batch processing texts performing tasks like tokenization, Named Entity Recognition, Part Of Speech Tagging, text classification, etc. *We do this without using third party APIs or managed services* This architecture generalizes to other deep learning applications as well.

- The following constraints of working with serverless computing platforms and approaches to work with them:
- Package size restrictions
- Memory restrictions
- Speed
- An automated build and deploy strategy
- Tips and tricks to maximize performance

Developing locally and then deploying to the cloud can be tedious. This talk is about our experiences deploying a new Java EE microservice in our stock trader application, and how we integrated Apache Kafka, Kubernetes Secrets, Helm Charts, Docker workarounds, and Service brokers to make it easier.

A typical microservice flow involves at least two services interacting with each other via rest endpoints, and persisting that data in a database. Apache Kafka provides an alternative approach allowing for a Cluster to intact with Streams. Producers ‘publish', and Consumers ‘subscribe' to these streams. Apache Kafka is typically used in building real-time applications to reliably get and transform data from streams. Using a Stock Trading application , resembling most stock trading apps online, we will show how we integrated Apache Kafka, and how we were able to produce events from a Portfolio microservice, consume them in a Trade-History microservice, and persist that data in Mongo DB. This application is containerized with Docker, and deployed to a cloud environment using Kubernetes and Helm. We will focus on some of the techniques we used regarding Kubernetes Secrets, Shared Environment Variables, docker tricks, and methods and technologies we found that helped us develop and test locally and deploy just as easily with minimal changes.