[Alex R (alexr17)]

Why autoscaling broke: #104 added the wait-for-brainstore init container, enabled by default. Because the metric endpoint used a numeric port, GKE attributed the API's metrics to that init container, so the HPA had no API metrics to scale on.

With a numeric endpoint, GKE's pod discovery can attribute API metrics to a completed init container, leaving the HPA without usable event-loop metrics.
Declare api-metrics on the API container and reference that name in AutoscalingMetric, so GKE keeps only that port's discovery target.

Configuration	Expected behavior
GKE autoscaling enabled; default or custom port	Named endpoint resolves the configured API metrics port
Default, ingest, background, or renamed pool	Each metric definition selects its pool's API port
Startup gate enabled or disabled; sidecar present	API gauges retain the API container identity
Autoscaling disabled on AWS, Azure, or GKE	No autoscaling metric definition or additional metrics port
Autoscaling enabled on AWS or Azure	Existing unsupported-provider validation applies
Release/custom image tags; reused or partial values	No new values or image-version checks

Local collector autoscaling-0.2.2-gke.0	Gauge container	Port metadata
Numeric endpoint + completed startup gate	wait-for-brainstore	Missing
Named endpoint + completed startup gate + sidecar	api	8001
Named endpoint + sidecar	api	8001

[chatgpt-codex-connector]

Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-06T22:53:35.125764Z	ad1b845	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "Codex (@codex) review" or "Codex (@codex) security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.