Field Notes

Field Notes are production-style conversations for Kubernetes LLM platform teams. Each note starts from a failure mode that looks simple from a dashboard but becomes a platform decision once traffic, GPU capacity, tenant boundaries, and rollout controls are involved.

Use these notes before design reviews, launch readiness checks, and incident retrospectives.

Field note	Production guide	Matching lab
LLM Latency War Room	LLM Latency on Kubernetes	vLLM inference challenge
GPU Capacity Incident	GPU Node Pool Scheduling for LLM Inference	GPU node pool scheduling
RAG Tenant Isolation Review	RAG Tenant Isolation on Kubernetes	RAG retrieval challenge
KServe vs Ray Serve Ownership	KServe vs Ray Serve for LLM Platforms	KServe vs Ray Serve decision

How to read a field note

Each note uses the same review shape:

1. Start with the visible symptoms.
2. Name the common wrong instinct.
3. Split the system into platform layers.
4. Pick the signals that prove or disprove the theory.
5. Decide what should be automated, documented, or moved into a lab check.

Why this section exists

Most Kubernetes LLM failures do not start as tool-selection problems. They start as ownership problems: who owns latency, GPU placement, retrieval quality, runtime flags, rollback boundaries, and cost evidence. Field Notes make those conversations explicit before a platform team scales the model.

Related pages

• K8s LLM: Kubernetes LLM Platform Guide
• Production Guides
• LLM Production Readiness Checklist
• LLM on Kubernetes
• Kubernetes LLM Labs
• Reference Architectures