roachprod: Add support for ASG groups in AWS

[cpj2195]

This PR adds AWS Auto Scaling Group (ASG) support for roachprod, mirroring the existing GCE Managed Instance Group (MIG) implementation. Managed clusters enable dynamic cluster resizing via roachprod grow and roachprod shrink
operations.

Summary

Add --aws-managed flag to create clusters using Launch Templates and Auto Scaling Groups
Implement Grow() and Shrink() operations for managed AWS clusters
Implement proper cleanup of ASG, launch templates, and load balancer resources on cluster destroy
AWS ASG doesn't support creating instances with custom names (unlike GCE MIG which has create-instance --instance ). To preserve roachprod's naming conventions (e.g., clustername-0001, clustername-0002), we:

Create the launch template (stores instance configuration)
Create the ASG with desiredCapacity=0 (management structure only)
Create instances using runInstance() with proper names
Attach instances to the ASG using attach-instances
This approach maintains compatibility with roachprod's cluster naming while gaining ASG benefits for grow/shrink operations.

For managed clusters, target groups are attached directly to the ASG using attach-load-balancer-target-groups. This provides:

Automatic instance registration: New instances launched or attached to the ASG are automatically registered with the target group
Automatic deregistration: Instances removed from the ASG are automatically deregistered
Simplified grow operations: No need for manual register-targets calls when adding nodes
Resource Naming

Launch Template: {cluster}-lt
Auto Scaling Group: {cluster}-{region}-asg
Epic: None
Fixes: #153071

[blathers-crl]

Your pull request contains more than 1000 changes. It is strongly encouraged to split big PRs into smaller chunks.

_{🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf.}

[cockroach-teamcity]

This change is 

[github-actions]

Potential Bug(s) Detected

The three-stage Claude Code analysis has identified potential bug(s) in this PR that may warrant investigation.

Next Steps:
Please review the detailed findings in the workflow run.

Note: When viewing the workflow output, scroll to the bottom to find the Final Analysis Summary.

After you review the findings, please tag the issue as follows:

• If the detected issue is real or was helpful in any way, please tag the issue with O-AI-Review-Real-Issue-Found
• If the detected issue was not helpful in any way, please tag the issue with O-AI-Review-Not-Helpful

[nameisbhaskar]

Thanks for the change. As discussed, lets make it to smaller modular functions.

[github-actions]

Potential Bug(s) Detected

The three-stage Claude Code analysis has identified potential bug(s) in this PR that may warrant investigation.

Next Steps:
Please review the detailed findings in the workflow run.

Note: When viewing the workflow output, scroll to the bottom to find the Final Analysis Summary.

After you review the findings, please tag the issue as follows:

• If the detected issue is real or was helpful in any way, please tag the issue with O-AI-Review-Real-Issue-Found
• If the detected issue was not helpful in any way, please tag the issue with O-AI-Review-Not-Helpful

[nameisbhaskar]

LGTM! Please do test the flows, including failure flows. Thanks!

[golgeek]

Left a few comments here and there, but great job!

I wonder how much more time it takes to spin up a managed cluster vs a non-managed cluster?
Idea behind this question is around making grow and shrink something that could be an afterthought as in you provisioned a non-managed cluster, but would like to grow it later on.
I assume that with the creation of the launch template, ASG and attaching the instances, it is much slower than with a regular cluster, and we shouldn't go that route?

That being said, with the way you implemented managed clusters, we could probably be able to convert a regular cluster to a managed one later on?

In the same kind of general direction, what would be your thoughts about using Unmanaged Instance Groups instead of MIGs in GCP?

[herkolategan]

Good work! Did a quick pass, nothing jumped out at me, that @golgeek hasn't already mentioned.

[github-actions]

Potential Bug(s) Detected

The three-stage Claude Code analysis has identified potential bug(s) in this PR that may warrant investigation.

Next Steps:
Please review the detailed findings in the workflow run.

Note: When viewing the workflow output, scroll to the bottom to find the Final Analysis Summary.

After you review the findings, please tag the issue as follows:

• If the detected issue is real or was helpful in any way, please tag the issue with O-AI-Review-Real-Issue-Found
• If the detected issue was not helpful in any way, please tag the issue with O-AI-Review-Not-Helpful

[cpj2195]

bors r+

[craig]

Build succeeded:

• unit_tests
• linux_amd64_build
• lint
• acceptance
• check_generated_code
• local_roachtest_fips
• docker_image_amd64
• examples_orms
• linux_amd64_fips_build
• local_roachtest