feat(aws): add elastic beanstalk environment

[lukeadams84]

Objective:

Add support for aws_elastic_beanstalk_environment and aws_elastic_beanstalk_application as FREE. Fixes #1410.

Pricing details:

Price breakdown for elastic beanstalk environments is ec2 instances associated with the environment multiplied by the minimum ASG numbers.

Additional costs are associated for the load balancer type. If StreamLogs is enabled, there can be usage costs for cloudwatch metrics, including ingest, scanning and storage in GB.

Status:

• Generated the resource files
• Updated the internal/resources file
• Updated the internal/provider/terraform/.../resources file
• Added usage parameters to infracost-usage-example.yml
• Added test cases without usage-file
• Added test cases with usage-file
• Compared test case output to cloud cost calculator
• Created a PR to update "Supported Resources" in the docs

Issues:

None

[hugorut]

@lukeadams84 thanks for doing this ❤️ . We've got some big planning meetings today so I might not get to review/look over this today, but i'll endeavour to get back to you as quick as I can.

[hugorut]

@lukeadams84 thanks again for this. I've had a peek through the code and done some reading. I'm wasn't familiar with elastic beanstalk environments so I appreciate you bringing it to our attention. My initial thoughts though is that we might need to go a different route with this resource than the norm.

My understanding that elastic beanstalk environment essentially simplifies provisioning a set of underlying AWS resources to run an application. In terraform you can change the base set of configuration options for this using the setting block. Because a lot of these settings are resources that Infracost already supports (ec2 instances, autoscaling group, cloudwatch, launch configuration, e.t.c.) my preference would be to use this resource fetch the configuration and then delegate pricing to those Infracost resources.

This would mean that all aws.ElasticBeanstalkEnvironment would do is call other resources like aws.Instance to output their cost components to the CLI. This would be more maintainable in the longer term as it would mitigate code drift between resources that are, under the hood, the same.

Let me consult with the team and get back to you.

[lukeadams84]

@lukeadams84 thanks again for this. I've had a peek through the code and done some reading. I'm wasn't familiar with elastic beanstalk environments so I appreciate you bringing it to our attention. My initial thoughts though is that we might need to go a different route with this resource than the norm.

My understanding that elastic beanstalk environment essentially simplifies provisioning a set of underlying AWS resources to run an application. In terraform you can change the base set of configuration options for this using the setting block. Because a lot of these settings are resources that Infracost already supports (ec2 instances, autoscaling group, cloudwatch, launch configuration, e.t.c.) my preference would be to use this resource fetch the configuration and then delegate pricing to those Infracost resources.

This would mean that all aws.ElasticBeanstalkEnvironment would do is call other resources like aws.Instance to output their cost components to the CLI. This would be more maintainable in the longer term as it would mitigate code drift between resources that are, under the hood, the same.

Let me consult with the team and get back to you.

Yes, I understand. I think one of the headaches seems to be that the EC2 instances (for example) that elastic beanstalk uses are more constrained than if you were creating an EC2 instance normally. I believe the instances are always dedicated, and only support Linux as an operating system.

I'll have another look at your examples and see what I can work out to use the existing resources.

[hugorut]

I'd recommend having a peek at eks node group, launch template or neptune cluster snapshot as they do similiar-ish things

[lukeadams84]

I'd recommend having a peek at eks node group, launch template or neptune cluster snapshot as they do similiar-ish things

Okay, have updated now. It looks like it covers the bases.

[hugorut]

@lukeadams84 amazing - I'll take a peek later

[hugorut]

Starting to shape up nicely. There's a few points to resolve before we get this in though.

[lukeadams84]

Can you tell I am new to go? :D Ill get on these, thanks for looking at it.

[hugorut]

@lukeadams84 not at all! It's just a bit of a bizarre resource, you chose a tricky one to start! Ping here if you have any other questions. Also feel free to message me on slack if you have any ad hoc questions that you want to run by me with a quicker response rate

[hugorut]

@lukeadams84 we're getting closer, just need to fix up these usage params.

[hugorut]

Minor code tweaks, but logic looks solid

[hugorut]

LGTM 🚀

[hugorut]

@lukeadams84