Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

As you're guessing that Cloudfare superior JS runtime plays a big role, it could be interesting to see if it can compete against Golang Lambda as well.

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#5

Post by jedberg » Mon, Jul 02, 2018, 9:59 PM UTC

The value of any function-as-a-service is the ecosystem within which it sits. Pretty much all of them are the same: upload your code, we will run it.

The value comes from 1) What can trigger that code to run and 2) What services that code can interact with.

And on those two points, AWS still wins hands down. They have by far the most possible triggers for Lambda, and they have by far the most services that Lambda can interact with.

It's cool that Cloudflare built something faster, but unless you're running in a vacuum, speed is the least of your concerns.

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#6

Post by mrkurt » Mon, Jul 02, 2018, 10:01 PM UTC

Comparing Workers to Lambda proper seems silly. Lambda lets you connect to DBs, use a lot more than 128MB of memory, etc, etc, etc.

Comparing them to Lambda@Edge makes sense, but Lambda@Edge is not a very good product.

(Full disclosure: my company competes with Cloud Flare Workers).

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#7

Post by zackbloom » Mon, Jul 02, 2018, 10:03 PM UTC

poulpi wrote: post #4

Have you think about including Golang based Lambda function in your benchmark? As you're guessing that Cloudfare superior JS runtime plays a big role, it could be interesting to see if it can compete against Golang Lambda as well.

I have a test which dives into the crypto performance, which seems to be largely driven by the amount of CPU allocated to the process (both Workers and Lambda is ultimately just calling a C crypto implementation). I'll have a longer post about it shortly, but the summary is a 128MB Lambda is around 8x slower than a Worker in pure CPU.

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#8

Post by kevan » Mon, Jul 02, 2018, 10:04 PM UTC

>To be fair, comparing my Lambda, which only runs in us-east-1 (Northern Virginia, USA), to a global service like Workers is a a little unfair.

At least you acknowledge that it's a bit silly to use a global benchmark to compare a global service with an intentionally-regionalized service.

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#9

Post by zackbloom » Mon, Jul 02, 2018, 10:06 PM UTC

kevan wrote: post #8

>To be fair, comparing my Lambda, which only runs in us-east-1 (Northern Virginia, USA), to a global service like Workers is a a little unfair. At least you acknowledge that it's a bit silly to use a global benchmark to compare a global service with an intentionally-regionalized service.

Why would you run something in a single location if you can run it everywhere for the same price though? It's not like Lambda is cheaper for being centralized.

Re: Serverless Performance: Cloudflare Workers, Lambda and LambdaEdge

#10

Post by zackbloom » Mon, Jul 02, 2018, 10:11 PM UTC

jedberg wrote: post #5

The value of any function-as-a-service is the ecosystem within which it sits. Pretty much all of them are the same: upload your code, we will run it. The value comes from 1) What can trigger that code to run and 2) What services that code can interact with. And on those two points, AWS still wins hands down. They have by far the most possible triggers for Lambda, and they have by far the most services that Lambda can…

Yes, that's how Amazon creates lock-in. But it depends what you're doing with it right? If you are looking to run code based on a SQS event, yes you have to use a Lambda. If you are looking to execute code when something visits a URL you have more options.