Using Proget and SynchroFeed to drive componentization and automation
At Alkami, we have been componentizing our monolith application by breaking it down into packages. These packages are nuget packages for .NET libraries and choco packages for microservices and widgets. Proget is our package server which is fed artifacts from our Bamboo server. We then use a product we open sourced called SynchroFeed to validate packages and mirror packages to our external feeds. This presentation will highlight Alkami's CI/CD architecture using Bitbucket -> Bamboo -> Proget -> Whitesource -> SynchroFeed -> Team City for deployment and how it allowed us to handle the explosive growth in microservices.
Outline/Structure of the Presentation
- About Me
- We started with a monolith...and life was good
- Unfortunately, the monolith was getting messy - Too big, Too much testing, Too slow to deploy
- We need to componentize - The march to microservices
- Challenges breaking down the monolith
- Alkami's CI/CD Toolset
- Pre-release build process
- Pre-release validation process
- Whitesource - Security and Compliance
- SynchroFeed Open Source Project
- What can I do with SynchroFeed?
- SynchroFeed Command Pipeline - Validate and Catalog Package
- SynchroFeed and Slack - Communicate failures to Development team
- Pre-release/Release Deployment - Team QA testing
- Release Deployment - Integrated QA testing
- Release Deployment - Staging Deployment
- Release Deployment - Production Deployment
- SynchroFeed External Feed Synchronization - Supporting our SDK Customers
- SynchroFeed Proget sync to folder - reducing load on Proget server
How Alkami handled componentization and the move to microservices which caused an explosion of growth in repos, build plans, packages and deployments.
Developers, DevOps, System Reliability Engineers
Prerequisites for Attendees
The participants should have an interested in automating CI/CD through packages.
schedule Submitted 4 years ago
People who liked this proposal, also liked:
Alex Papadimoulis - Opening Remarks: Choosing Correctly in an adapt-or-die worldAlex PapadimoulisPresident and FounderInedo
schedule 4 years agoSold Out!
Simplicity is dead. At least, that’s what never-ending stream of constantly changing software platforms would lead you to believe.
Should you use function-based cloud computing? Platform-as-a-service? What about containers, or is it container orchestration? And this is all before you even get to the software itself, and the layers upon layers of abstractions and packages that need to be considered before even thinking about writing code that serves the business.
We’re in the middle of development technology paradigm shift, living in an adapt-or-die world. But there are a lot of key decisions to make, and I’ve seen a lot of organizations make the wrong decisions and end up down a pretty grave path.
Ultimately, they end up with mission-critical systems that are too brittle to change, and too broken not to. The clever engineers that design the systems eventually leave, and no one else is willing to fix the mess. What seemed like a nice technological win just ended up leaving the organization struggling to maintain the systems, while competitors run laps with better, easier-to-maintain systems.
In this talk, I’ll show how, by continuing to partner with Inedo, you can not only avoid becoming a casualty of this paradigm shift, but how you can out-innovate and demonstrate to the powers-that-be that your software platform will let them run laps around their competitors.
I’ll also talk about my vision for Inedo’s future, including our product roadmap and how we’ll continue helping organizations address constantly emerging technology.
Sneak peek: it’s all about getting back simplicity and creating real value by building software