travisgosselin.comTravis J. Gosselin | Software Developer, Architect & Speaker

Travis J. Gosselin | Software Developer, Architect & Speaker Travis J. Gosselin Software Developer, Architect & Speaker TOP ABOUT SKILLS &AMP INTERESTS TALK ABSTRACTS BLOG EMPLOYMENT EDUCATION CONTACT ABOUT ME a little about me. &OpenCurlyDoubleQuote I do not fear computers, I fear the lack of them! (Asimov) &CloseCurlyDoubleQuote Travis is an accomplished software developer, architect, and periodic speaker. A tech enthusiast and blogger, Travis finds his niche in architecting and working with teams to compose highly automated service-oriented systems both in the cloud and on-premise. Travis currently works as a Principal Software Engineer for SPS Commerce, the world's largest retail network. He has a strong focus on continually learning new architectures and development patterns across different stacks and technologies, but has a special place in his heart for DevOps and making the software development lifecycle (SDLC) more efficient. Name Travis Gosselin Home Toronto, Canada Twitter @travisjgosselin GitHub travisgosselin SKILLS & INTERESTS what i'm interested in. Development .NET & .NET Core Java Node, TypeScript NPM, Angular, React Relational Databases, NoSQL, Queues & Streams Architecture API Design Synchronous & Asynchronous Workflows DevOps, CI/CD, 12 Factor App, Automation Bounded Contexts, Data Domains & Microservices Solution Architecture ABSTRACTS presentation and speaking topics Unleashing Deploy Velocity with Feature Flags A lot of development teams have built out fully automated CI/CD pipelines to deliver code to production fast! Then you quickly discover that the new bottleneck in delivering features is their existence in long-lived feature branches and no true CI is actually happening. This problem compounds as you start spinning up microservices and building features across your multi-repo architecture and coordinating some ultra-fancy release schedule so it all deploys together. Feature flags provide you the mechanism to reclaim control of the release of your features and get back to short-lived branches with true CI. However, what your not told about feature flags in those simple "if/else" getting started demos is that there is an upfront cost to your development time, additional complexities and some pitfalls to be careful of as you begin expanding feature flag usage to the organization. If you know how to navigate these complexities you will start to unleash true velocity across your teams. In this talk, we'll get started with some of the feature flagging basics before quickly moving into some practical feature flagging examples that demonstrate its usage beyond the basic scenarios as we talk about UI, API, operations, migrations, and experimentation. We will explore some of the hard questions around "architecting feature flags" for your organization. Microservice Code Sharing with .NET Service Templates Sharing code and internal libraries across your distributed microservice ecosystem feels like a recipe for disaster! After all, you have always been told and likely witnessed how this type of coupling can add much friction to a world that is built for high velocity. But I'm also willing to bet you have experienced the opposite side effects of dealing with dozens of services that have had the same chunks of code copied and pasted over and over again, and now you need to make a standardized, simple header change to all services across your platform; talk about tedious, frictional, error-prone work. This is where service templates come in! A service template provides immense power to roll out changes that affect each service from a single code base. When built correctly, such a template stays far away from sharing any domain objects or business rules and focuses on delivering appropriate coupling between your services that allows your team to benefit from technical and operational component updates. In this talk, we'll explore the architectural myth in microservices that you should NEVER share any code and explore the dos and don'ts of the types of reuse that you want to achieve through appropriate coupling. We will dive deep and look at some code of what a Service Template architecture looks like specifically in .NET, while also spending some time on the life cycle of that template and how it rolls out effectively to all your services. We will finish it off with some considerations and pain-points you might encounter with service templates in the organization. Mono-repo or Multi-repo? Do I Need to Choose Just One? Over the last few years, the dichotomy over Mono-repo vs Multi-repo approaches to source control has trended in the tech community and now heavily in the organization. If large organizations like Google and Facebook have been using mono-repos for a decade then shouldn't I be using them too? Like everything in software architecture, the answer is: "it depends!". While there are significant advantages to mono-repo style source control, there are significant disadvantages as well. Some developers swear by one approach over another, but in reality, you probably have different architectures that naturally would benefit from each style. Understanding when to use each approach over another is key to ensuring that you don't end up continually expanding and contracting your repository granularity without good reasons. In this talk, we will move through a strong comparison of a mono-repo vs multi-repo approach. Each one represents a different set of values that may more closely align with your team or organization. We will cover the trade-offs, some workarounds and architectural considerations for helping you choose the correct style for your teams and projects. Getting Started with Azure Pipelines Azure Pipelines is a DevOps cloud-hosted tool from Microsoft that enables you to build efficient and streamlined CI/CD build and deployment pipelines for just about anything. Contrary to its name, it is cloud-agnostic and can provide your small or large team with a very flexible and inexpensive way of delivering your software with a high degree of consistency. In this demo driven intro talk, we will dive into a brief overview of Azure DevOps and use Pipelines to build a CI/CD pipeline to deploy a service to Azure Cloud from GitHub. We will evolve our pipeline as we add in immutable versions, blue/green style deploys, containers, YAML, and make sure we get all the Pull Request status checks integrated as well. Then we will quickly convert it over to deploy into AWS! BLOG records of a software developer. .NET Core and AWS Secrets Posted by: travis No Comments .NET Core with runtime secrets has been a bit of a journey within the AWS ecosystem over the past 5 years. Our journey onboarding and shifting our entire infrastructure from on-premise into AWS Cloud started in 2016. I know I'm... Read More Moving from Dependabot-Preview to Dependabot Native Posted by: travis No Comments Dependabot native has been around for a couple years now after GitHub officially acquired it in 2019. But if I google "Dependabot" I still generally find myself at the "Dependabot.com" home-page, and up until last week found myself still using... Read More AWS Cross-Account Resource Access Patterns Posted by: travis 1 Comment AWS Account Granularity The growth of SPS Commerce has continued to be very strong, even amidst the recent global pandemic, as we work to provide an enabler for essential services. The demand for SPS services and products continues to grow... Read More Meetup: DevOps Underground London Posted by: travis No Comments Had a great opportunity to join up with DevOps Underground Meetup from London today to talk more about Feature Flags after a great talk on Microservices. It's always refreshing to talk with the community of engineers, and while feedback is... Read More Architectural Unit Testing Posted by: travis No Comments I'm going to assume that you already buy into the advantages of unit testing your code, ...