This week’s serverless superhero is Jye Cusch, co-founder and CEO of Nitric, a company focusing on infrastructure from code. Jye is one of the few people pioneering the IfC space and does a great job sharing his adventure with us on the exciting stuff he’s building. He’s a great person to talk to if you want to get insights on where serverless engineering is headed. Thank you Jye for all your hard work and innovation!
Kieran Wrenn starts us off with an interesting article on our options for custom domain names with microservices. He talks about the different strategies you can employ to balance tradeoffs between a big blast radius and having disparate domain names. Tip to readers, he accidentally uses the word “endpoint” instead of “API” throughout the article which might lead to some confusion. Go into it knowing he means API and it reads very well.
In part two of his series on data caching, Alex Debrie shares with us 6 common caching design patterns to execute your caching strategy. In this post Alex talks about the different caching implementations and the pros and cons of each.
Best practices help us normalize on convention in the wild west we know as “the cloud”. That is why when we see content from Ran Isenberg we need to stop what we’re doing and look at it immediately. He recently shared with us some best practices on how to deploy to AWS with GitHub Actions and AWS CDK. He provides a code sample plus some details on how to store secrets and when to run your different types of tests.
Observability is often talked about with serverless applications for good reason. It is critical to success in any production scenario. Siben Nayak walks us through observability in distributed systems in great detail. He covers the different pillars of observability, types of implementation, different tools you can use, and even tells us about best practices. This might be one of the most thorough observability articles I’ve read that wasn’t an advertisement. Great stuff!
The race to get to the cloud is…. interesting. You have “lift and shift” migrations that take apps out of data centers and run them in the cloud on VMs. You also have “purists” who will never look at anything that isn’t serverless. But is that really what the future of compute is all about? Jeremy Daly talks to us a bit on his thoughts of cloud native vs native cloud apps, discussing some fallacies of moving to the cloud and how the future holds a balance of everything we know today. It’s an insightful post from a visionary in the modern software world.
A new Lambda extension was launched last week enabling faster access to SSM Parameters and Secrets Manager secrets. This is a huge update for performance considerations when accessing secrets. In our last issue we even had an article featuring how to secure secrets in a Lambda function. This will take that recommendation to the next level.
A big announcement for Step Functions was made last week. Express workflows now have multiple visualization options. This allows you to view them the same way you would a standard workflow, which was my primary detractor from using them. Fantastic update!
Developers rejoice as AWS announces dark mode support for the management console. This definitely isn’t a serverless announcement, but is a nice update if you’re a dark mode user.
Last week was DataDog Dash, their annual conference around building and scaling the next generation of applications. In case you missed it, they announced some incredible new features. Some of the new features include CoScreen - a remote screen sharing feature for streamlined debugging, Stream Monitoring - a new way to view event-driven pipeline health in real time, and an Intelligent Test Runner - a tool to automatically select and run affected tests as a result of a code change. There are many more new features that will enable invaluable insights in our applications.
Whether we admit it freely or not, we all secretly know that great observability is the key to success with serverless applications. Khawaja Shams shared a thread on lessons he learned from running services like DynamoDB.
The worst alarm for a #serverless service is customer reachout. If customers detect a problem before us, we’ve failed.— Khawaja Shams (@ksshams) October 19, 2022
An 🚨alarming thread🧵 on alarms based on lessons @akshatvig & I learned from running services like #dynamodb.
Please add your favorite lessons inline /RT.
If you haven’t registered for it yet, the Serverless Summit is next month. It is a free event featuring rockstar engineers from the serverless community.