Does lambda memory allocation impact network bandwidth?

Question

Even though the CPU and Memory is < 30% utilization at 1024mb, it appears that changing the lambda to 2048mb doubles the network throughput. My runtime is single threaded node.js so it shouldn't particularly be impacted by the additional vCPU. I can't find specific details about network scaling, it's just vaguely hinted at:

https://docs.aws.amazon.com/lambda/latest/operatorguide/computing-power.html

> Adding more memory proportionally increases the amount of CPU, increasing the overall computational power available. If a function is CPU-, network- or memory-bound, then changing the memory setting can dramatically improve its performance.

Is there something more concrete to refer to?

Accepted Answer

Yes, the network performance will increase with memory. There are no specifics as there is no definitive performance since Lambda compute is abstracted from you and it is also protocol and application specific. You should benchmark your function with various memory sized to determine where you meet your requirements.

Answer

If your lambda is interacting with resources inside your VPC or with AWS services, you should VPC enable your lambda function if you haven't done so already - https://docs.aws.amazon.com/lambda/latest/dg/configuration-vpc.html

Also make use of interface VPC endpoints for AWS services. That way the traffic goes over the AWS network and not the internet - https://docs.aws.amazon.com/vpc/latest/privatelink/create-interface-endpoint.html

Answer

All Lambda functions get the same type of network interface with the same bandwidth. However, the more memory you give your function, the more CPU it gets, which allows it to actually drive the network bandwidth, up to the limit that the network interface supports. This is why you will see an increase in bandwidth only until some point.

Does lambda memory allocation impact network bandwidth?

Relevant content