DEV Community

Discussion on: A crash course on Serverless with AWS: Centralized logging with Kinesis and Lambda

 
adnanrahic profile image
Adnan Rahić

Let me jump in. The idea isn't to choose something over something else. You choose the best tool for the job.

Here's an example. You want to resize an image on the fly, once it's uploaded to your S3 bucket. Configure an async event trigger from S3 to run a Lambda function, resize the image and put it back into the S3 bucket. This is done automagically, without you having to worry about it. It's better to be configured with a serverless architecture than a conventional server.

Another example. You have an API with huge I/O operations with heavy throughput and many database connections. You'd want a fast server running a language like Go to handle the concurrency. Having a traditional setup is superior in this case.

Do you get in what way we should be moving? Don't bring a shovel to a knife fight. :)