This week our serverless superhero is Ryan Jones, founder of Serverless Guru. Ryan is constantly involved in serverless conferences, panel discussions, podcasts, and so much more. Ryan lives and breathes serverless, and has done an amazing job helping to shape the community and build a team who genuinely wants to help people succeed with it. Thank you Ryan, for all that you do!
Lee Gilmore starts us off with an insightful and incredibly in-depth post on building the experience layer of your serverless application. He walks us through what it is, how it relates to serverless, and what apps would be like without it. This is the first post of a series on the various layers of a serverless enterprise application, so stay tuned!
A question regularly asked when discussing event-driven architectures is “how do you troubleshoot errors?” Well, step one is to make sure you log all the events flying back and forth in your application. Robert Slootjes shows us how to capture all EventBridge events and push them into S3 via a Kinesis firehose. His post also includes how to use the CDK to format the events for easy downstream parsing.
If you’ve ever felt stuck trying to debug issues with an API Gateway integration, you’re not alone. Arpad Toth points us in the right direction with his post on how to debug generic API Gateway errors with access logs. He demonstrates how to turn them on and what to log to get to the solution the quickest.
Step Functions recently added 14 new intrinsic functions. Last week Ben Smith gives us some reasons why we should be excited about them. They speed up execution, lower costs, and provide less code to maintain. He gives us some real world examples of how to update your workflows to utilize the new functions.
An update came to DynamoDB last week allowing for up to 100 actions per transaction. This is one of those updates where I think just because you could doesn’t mean you should. Hopefully you don’t have a real use case for 100 actions in a single transaction in your serverless app. If you do, it might be time to revisit the data model. Nonetheless, if it gets people to the cloud faster with their legacy systems, then I’m happy with it.
Have you ever had a conversation with someone about serverless and they say “but what about cold starts”? Yeah, me too. Cold starts seem to be feared by many, but not many people know why. I (Allen Helton) wrote an article suggesting we stop talking about cold starts. The topic is rarely productive and they are mostly a solved problem or a non-issue.
There was a great high-level blog post from McDonald’s on how they are building event-driven architectures. David Boyne sums up the highlights for us in a thread.
Just been reading how McDonaldβs are doing event-driven architectures ππ
— David Boyne π (@boyney123) September 9, 2022
β οΈ Importance of standards
π Schema registry + validation
π Event Gateway Pattern
π§βπ» Custom SDKs
β Defined flow of events
Here are some thoughts (with blog links) π§΅π
If you weren’t able to attend, the content from EDA Day will start appearing online soon. Once it does, I’ll be sure to highlight the material in the newsletter. From what I hear, it was a stellar conference with fantastic speakers and content.
If you’d like to make a recommendation for the serverless superhero or for an article you found especially useful, send me a message on Twitter, LinkedIn, or email.
Happy Coding!
Allen
Thank you for subscribing!
View past issues.