This years AWS Re:Invent was a great event. Nice talks, nice talking and a lot of amazing new services announced for AWS. One of them is AWS Lambda, a nodejs based computation service which adds a lot more abstraction of computation from resources than ec2 does. You don’t have to care about instances, software or resources. Just define a function you wan’t to execute and an event triggering it. That’s it so far. This sounds trivial and I don’t want to boast it’s meaning but it can be a great solution if you wan’t to wire some AWS services and do trivial transformation or aggregation on processed data. Imagine a kinesis stream where ELB log files flow through. You can define a lambda function gathering new log-files from s3 log buckets and another function which counts events in the log stream and stores them in dynamodb for real-time dashboards. All of this without the need to run any ec2 instance with all the complexity it brings if it has to be done the right way.
The service is in a public beta phase, register here to participate: https://aws.amazon.com/de/lambda/
Documentation is available here: https://docs.aws.amazon.com/lambda/latest/dg/lambda-introduction.html