4 Answers
Hi, I’m not sure which lesson you heard this in, but "bake all your dependencies" most likely refers to creating a custom AMI in which you install all the software that you need to run your application and saving it as a custom AMI.
This appears as an answer in the High Availability and scaling Quiz.
The term ‘baked in’ means as Marc said above.
I agree with what was said, but I’d like to add as an example, when you write a python (or most programming languages) sometimes you need to include other libraries to get a certain function… for example zip() is a function in the standard python library, but pyping() is a library you have to go to an external library (PyPy I think) to include at runtime… You could also use a boot-time script to add dependencies to an EC2 instance… those dependencies being Services, like httpd, or auditd(). With Docker containers you load a program into a container, and in that container are included all the dependencies for your app, which is similar to a stripped down OS, but there is no kernel required in a Docker container.
You would want to do this when your warm up times for EC2 instances hinder your load balancer from scaling up in a reasonable time. This will reduce long instance warm up times, at the expense of having to maintain your own AMI. As mentioned above baking in is simply a baseline build with the required patches, services, libraries and applications.
Makes perfect sense given the context it was used.