Lambda Isn’t Made for Parallelism — But Go Still Gets the Job Done

Goroutines let you run multiple operations in parallel within one request. By moving blocking, I/O bound tasks into goroutines, you can unlock performance improvements by making operations concurrent.

Go is well known for how it supports and models concurrency in its ecosystem. Providing constructs like Channels, WaitGroups and Goroutines to build concurrent programs.

One thing that kept me intrigued was how this would work in a serverless system; in this case specifically AWS Lambdas.

• Your runtime is not long lived, lasts only as long as your function runs
• You don’t want background Goroutines either they’d get killed once the main handler returns
• Lets not forget the compute costs that add up the longer they run
• Core allocation when running lambdas is not transparent

Does this limitation really matter? Goroutines don’t directly map to OS threads, so you could still design your ...