fix EC2 placement zone retrieval [sc-180267]

[duncaan]

Current Behavior

service discovery of k8s pods does not work if you specify an AZ.

Why do we need this change?

We want to discover pods by their AZ.

Implementation Details

Dependencies (if any)

📇 [sc-180267]

[Copilot]

Pull request overview

This PR fixes EC2 placement zone retrieval for Kubernetes pods discovered through Consul by extracting the availability zone from the external-k8s-topology-zone metadata field. The change enables filtering and ordering of K8s pods by their availability zone, similar to EC2 instances.

Key changes:

• Updated map_k8s_consul_endpoint to extract availability zone from Service metadata
• Added test data and assertions to validate AZ retrieval from K8s pods
• Version bumped from 2.11.4 to 2.11.5

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 2 comments.

File	Description
lib/movable_ink/aws/ec2.rb	Modified map_k8s_consul_endpoint to populate placement.availability_zone from external-k8s-topology-zone metadata instead of nil
spec/ec2_spec.rb	Added external-k8s-topology-zone metadata to test data and replaced duplicate assertion with AZ validation
lib/movable_ink/version.rb	Bumped version to 2.11.5
Gemfile.lock	Updated MovableInkAWS version to 2.11.5

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The node_meta filter on line 132 filters based on Node.Meta.availability_zone, but for kubernetes pods, the availability zone is stored in Service.Meta['external-k8s-topology-zone']. This means when availability_zone parameter is provided, the Consul query will filter out all kubernetes pods (since they don't have the availability_zone in Node.Meta), but the code will still try to map them and set the placement zone from Service.Meta. Consider either: 1) filtering kubernetes endpoints in application code after retrieval, or 2) handling kubernetes and EC2 endpoints separately with different queries.