#101 — May 17, 2019

Read on the Web

Serverless Status
serverless news, views, and developments every Friday

What AWS Lambda’s Performance Stats Reveal — When you run an AWS Lambda tracing and logging service, you get to collate some interesting data. Amongst Epsagon’s userbase, 128MB and 1024MB are the most common memory configs, with Node 8.10 and Python 2.7 the most popular runtimes.

Ran Ribenzaft (Epsagon)

Upcoming Updates to the AWS Lambda Execution Environment — Rapidly approaching 5 years of age, the Lambda platform is undergoing a key update in terms of its underlying operating system. All new functions after May 21 will be on the new system, with existing functions upgraded on June 18, so you may want to start doing some testing soon.

Amazon Web Services

Get the Fastest Website Deployments. Get Started Free — Unlike most Continuous Integration (CI) solutions, Buddy only re-builds the parts of your application that have changed and deploys your updates in seconds to any cloud. Start building better Serverless apps faster with Buddy.

Buddy sponsor

AWS Lambda Adds Support for Node 10 — Node.js is central to many serverless platforms – indeed, it was AWS Lambda’s first supported runtime – and so it’s great to see the biggest one support a recent version of Node.

Amazon Web Services

An Application Study in Serverless and Azure with 'The Urlist' — Burke Holland Cecil Phillip, both of Microsoft, have built a site for creating and sharing lists of URLs which is also designed to be an exemplar of a serverless app running on Azure. It uses TypeScript and Vue.js on the frontend, C# in the back, and in this post they share how it’s structured. The code is on GitHub, naturally.

Burke Holland

The Lifecycle and Performance of a Lucet InstanceLucet is a native WebAssembly compiler and runtime that Fastly are working on with the aim of being able to rapidly execute WebAssembly on every HTTP request a server might receive.. such as for, say, a serverless platform 😄 This is a pretty technical look at how they get things so fast and how this could ultimately make serverless functions a lot faster soon.

Adam Foltzer (Fastly)

📝 Tutorials

Using AWS CloudWatch Insights to Fine Tune Lambda Functions — Every AWS service produces detailed logs and CloudWatch Insights provides fast, interactive analytics on those logs meaning you can analyze what your Lambda functions are doing too.

Kyle Galbraith

Building a Serverless Comments System — This isn’t how I’d do it but it makes me realize a commenting system is a great use case for serverless and there are lots of ways you could approach it. Here, the use of AWS AppSync, DynamoDB, and GraphQL are considered.

James Wren

Build Stateful Apps with Serverless & Postgres — Trigger Serverless functions on Postgres events & build resilient & scalable Serverless applications using the opensource Hasura GraphQL Engine.

Hasura sponsor

How to Control and Record Voice Calls with Node.js Serverless Functions — In case you forgot that communications platform Twilio has its own serverless functions platform, this well-written tutorial on using it to managing the recording of voice calls might whet your appetite.

Stefan Judis (Twilio)

▶  Serverless Audio Analysis with Google Cloud Functions and LibROSALibROSA is a Python package for audio analysis and it’s easily used from Google Cloud Functions.

Julian Rubisch

Lambda and Kinesis - Beware of Hot Streams — Has enhanced fanout for Kinesis streams resolved the problem of ‘hot streams’ in Lambda? Spoiler: A little, but not entirely.

Yan Cui

🐦 Seen on Twitter

“IMO, there’s nothing in infrastructure engineering that is currently more innovative than the serverless model of computing. From traditional FaaS to wasm on the edge, this is where we’re seeing real, bold innovation.”

___
Cindy Sridharan (@copyconstruct) commenting on Twitter about Fastly's post on Lucet.