---
title: "Cluster orchestration"
lesson_id: "11"
---

# Cluster orchestration

- **Lesson ID:** 11
- **Goal:** GPUs fail. Jobs get preempted. Storage is slower than you hoped. Write the scheduling and failure story before you scale.
- **Human lesson:** [11-cluster-orchestration.html](11-cluster-orchestration.html)
- **Phase:** Phase 6 Systems

## Prerequisites

- The previous week's artifact, or a written note if this is week 1.
- Python 3.11+ and a laptop. CUDA is useful from week 5 onward and not required to read.

## Inputs, outputs, and artifacts

- **Inputs:** The replica from the prior week.
- **Outputs:** A cluster runbook: queue, health, checkpoint, and what you do at 3am.
- **Artifacts:** Lesson notes plus the code named in the human page.

## Agent build steps

1. State the week 1 principle: you do not need to pretrain an 800B model to understand the shape of the problem; you do need progressively realistic replicas.
2. Teach the topics on the human page without inventing paper URLs or news claims.
3. Keep starter code in fenced blocks that match the human lesson.
4. End with the assignment and the opinion checkpoint.
5. Use Edge FDE tone: ship, do not sightseeing. Opinion checkpoints matter.

## Constraints

Plain spoken English. No quizzes. No em dashes. No invented metrics, dates, or citations. Brand Edge FDE only. Name well-known papers by title only: Attention Is All You Need, InstructGPT, Direct Preference Optimization, ZeRO, PagedAttention/vLLM, FlashAttention, Scaling Laws for Neural Language Models.

## Key concepts

- Orchestration is submit, place, run, checkpoint, reclaim, retry.
- Multi-GPU jobs need a fabric and a gang, not just more cards.
- Storage and quotas kill runs as often as kernels do.
- A runbook is part of the replica.

## Takeaways

- Write the four-page runbook.
- Name signals for dead ranks and full disks.
- Checkpoint cadence is a storage decision, not a superstition.

## Acceptance checks

- Human HTML has Key concepts and Takeaways.
- Opinion checkpoint is present.
- No em dash and no fake URL.
- [ ] Proceed to [lesson 12 brief](12-supply-chain-and-capstone.llms.md).