DEV Community

Cover image for State of Lambda Functions in 2020 by Dashbird
Taavi Rehemägi for Dashbird

Posted on • Edited on • Originally published at dashbird.io

State of Lambda Functions in 2020 by Dashbird

2020 - what a ride it's been! The state of the world this year has definitely been unpredictable and ever more surprising. So for some interesting insights and stability, we've turned to Lambdas. Namely, the state of Lambda functions in 2020! 

Ever wondered how others look without *their face masks on? At Dashbird, we get to (metaphorically speaking) see what everyone looks like behind their serverless masks (read: stacks)*, and identify trends and patterns. Our platform gives us the ability to look at the state of Lambda. Sharing is caring and we're happy to share these insights with you, offering the chance to compare and contrast your Lambda functions to others, or just have a nose around. 

For reference, here's our article on the state of Lambda functions in 2019.

What's a "normal" amount of Lambda functions? 

As we know, there is no "normal" but we see an average of 147 Lambda functions per client in total. A slight increase from our 2019 figures of 123 functions. 

average number of Lambda functions

What about their runtimes? 

Node.js is undoubtedly the most widely used, with Java8 and Python following behind. 

Lambda runtimes usage

And, what's the regional spread? 

There's no change here as us-east-1 and eu-west-1 remain the most popular, but it's great to know that over half of Dashbird accounts use more than one region. 

AWS Lambda Regions

Digging into the Functions 

Let's look at the functions and how they perform now. 

Average code size

Average memory size

Average timeouts

We can see that Java leads in all of these areas, which makes sense given its nature of a lengthier startup but also its smooth running once set up. 

X-Ray is being used with 16% of the functions on average

X-Ray usage

So, how did you compare? What we can say is, everyone has a nose and a mouth under their mask -- they often just look and feel a bit different from yours!

Top comments (0)