MSPE-391: Adding Demand Execution Constructs to CDK Lib

[rpmcginty]

• misc updates
• minor updates to constructs_.assets
• misc changes, import cleanup
• updating sfn fragments (data sync, demand)
• top-level stack changes to core app (example)

What's in this Change?

• adding fragments for demand execution step function
• providing example stacks for how to set it up
• fixing imports in various places

Testing

[njmei]

Would we be able to use this construct even for external buckets? Or is this only intended to be used in situations where we are creating (and fully controlling) the s3 bucket(s)?

[rpmcginty]

this is a simple construct and not something we would use in merscope specifically. but I would define the storage aibs s3 bucket and use similar to the buckets specified here.

[njmei]

Under what circumstances would we want to allow state_machine_name to be None? Aren't the autogenerated names really hard to track (when we need to track down where the state machine is defined)?

[rpmcginty]

yes, but sometimes we don't care exactly what it is. I am thinking about cases where the state machine will be a nested state machine from which we can find in the top-level named state machine.

[njmei]

minv_cpus (min_vcpus) still doesn't look to have been fixed in BatchEnvironmentConfig class?

[rpmcginty]

minv_cpus (min_vcpus) still doesn't look to have been fixed in BatchEnvironmentConfig class?

did you see my second response? I realized I specified it this way because this is how it is specified in the cdk library.

See here: https://docs.aws.amazon.com/cdk/api/v2/python/aws_cdk.aws_batch/ManagedEc2EcsComputeEnvironment.html#:~:text=will%20launch%20Instances.-,maxv_cpus,-(Union%5B