What's the difference between Elastic Beanstalk and Lambda?
In simple terms, with lambda you can ignore where your code is running. You deploy a package that AWS lambda will execute somewhere for you. You pay by execution time (you won't pay if you are not running any code for idle servers). There are some restrictions in terms of languages you can use and their environments, and deployments are a bit unorthodox. There are several projects trying to use lambda as the backend for a full web applications (in conjunction with API Gateway).
Elastic Beanstalk provides a more typical execution environment: you code you app as you'd usually do and AWS takes care of provisioning your web servers, app deployments, configuration management, logging and scaling among others. Basically you defer lots of ops work to them, but in the end you end up with a similar environment to the one you could do by hand with EC2, ASGs, ELBs, etc. Here as well you have some predefined environments you can use, much more varied than the ones allowed in lambda, and also Docker environments that will give you all the flexibility you might need.
In case you are wondering what you need between those two, the answer is most likely Elastic Beanstalk. Lambda is a really interesting project, but it's in its infancy and people are exploring ways to use it efficiently.
Elastic Beanstalk provides a more typical execution environment: you code you app as you'd usually do and AWS takes care of provisioning your web servers, app deployments, configuration management, logging and scaling among others. Basically you defer lots of ops work to them, but in the end you end up with a similar environment to the one you could do by hand with EC2, ASGs, ELBs, etc. Here as well you have some predefined environments you can use, much more varied than the ones allowed in lambda, and also Docker environments that will give you all the flexibility you might need.
In case you are wondering what you need between those two, the answer is most likely Elastic Beanstalk. Lambda is a really interesting project, but it's in its infancy and people are exploring ways to use it efficiently.