[rhanka]

The dev Elasticsearch container still exceeded its new 512Mi limit while loading the real data, even with ML disabled. Logs confirmed the requested 256m heap but also showed a 128Mi direct-memory ceiling and default roles for ML, transforms, remote clusters, and every data tier.

This keeps the requested 512Mi container limit and 256m heap, caps direct memory at 64m, and limits the dev node to the roles used by the application (master,data,ingest) and one Elasticsearch processor. The retained PVC and its data are unchanged.

Validation:

• exact container configuration started green with Docker --memory=512m
• Elasticsearch node API reported heap 256Mi, one allocated processor, and roles data,ingest,master
• JVM startup log reported MaxDirectMemorySize=64m
• kubectl kustomize deploy/k8s/overlays/dev
• kubectl apply --dry-run=client on the rendered overlay
• git diff --check
• harness check scope
• harness check branch

External review note: AGY Gemini 3.8 again exited immediately after initialization without returning a verdict. The complete diff was supplied without requiring tool access.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Advanced

Run ID: 207c067b-dbd8-4243-aa20-7236023489b2

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}