Skip to content

Conversation

@AlirezaShamsoshoara
Copy link
Member

Summary

This PR introduces a complete guide to Monarch Meshes, the organizational backbone for managing distributed systems in Monarch. The documentation covers:

  • Core Mesh Concepts: Detailed explanation of what meshes are, including their multidimensional structure, named dimensions, extents, and hierarchical organization
  • Three Mesh Types: In-depth coverage of HostMesh (compute hosts/machines), ProcMesh (OS processes), and ActorMesh (actor instances), including their characteristics, relationships, and usage patterns
  • Mesh Hierarchy: Visual diagrams and examples showing how meshes spawn other meshes, dimension inheritance, and the complete spawning flow from hosts to processes to actors
  • Mesh Operations: Comprehensive guide to slicing (single index, range, and multi-dimensional), broadcasting, point-to-point selection, and extent manipulation
  • Distributed Patterns: Four common distributed computing patterns with code examples and diagrams:
    • Simple Data Parallel
    • Parameter Server
    • Pipeline Parallel
    • Hierarchical Communication
  • Best Practices: Production-ready guidance on mesh design, slicing patterns, dimension naming, extent validation, resource management, and testing strategies
  • Visual Aids: 15+ Mermaid diagrams illustrating mesh structures, hierarchies, operations, and communication patterns

This comprehensive guide serves as the definitive resource for understanding and implementing mesh-based distributed computing in Monarch, complementing the existing actor and system overview documentation.

@meta-cla meta-cla bot added the CLA Signed This label is managed by the Meta Open Source bot. label Nov 18, 2025
@AlirezaShamsoshoara
Copy link
Member Author

cc @johnwhumphreys @colin2328 Adding you here since I cannot add reviewer directly

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Meta Open Source bot.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant