sig-large-scale-cluster
Welcome to sig-large-scale-cluster!
Working Goals and Scope
- Goals: Dedicated to providing stable ultra-large-scale clusters, focusing on Kubernetes core component optimization, AI job scheduling optimization, and TCP store keychain optimization in a single cluster. Breaking through Kubernetes' declared multi-cluster limits, supporting 128k card (1.6w node) clusters. Facing future 50w card cluster scale, providing multi-cluster multi-level scheduling and other technologies to further improve scheduling performance.
- Scope: Focusing on network, monitoring, AI job scheduling, image pull acceleration, AI training stability improvement, Kubernetes core component optimization, and automated cluster deployment in ultra-large-scale scenarios.
Organizational Regular Meetings
sig-large-scale-cluster is open. Our users, contributors, and maintainers continuously improve it. It is because of you that we can build a more stable and easy-to-use ultra-large-scale cluster. Join us!
We will share the latest progress, solve project problems, demonstrate new features, and discuss future plans in our meetings. Join the sig-large-scale-cluster regular meeting:
- Meeting time: Every Thursday morning 10:00~11:30.
- Meeting link: Get it in openFuyao official website community meeting.
- Meeting minutes: Meeting minutes.
Members
Maintainer List
- Luo Gangyi @luogangyi
- Zhu Guihua @osmcn
- SHI Zilong @shizilong
Contact Methods
- Mailing list: Subscribe to sig-large-scale-cluster email.
Project List
For repository addresses, please see code-repository-info file.
Directory Structure Description
- This repository directory structure description
| Directory Name | Description |
|---|---|
| best-practices/ | SIG internal best practice summary. |
| design/ | Finalized design documents, including architecture design, feature design, review results, etc. |
| discussions/ | Repository for non-formal, exploratory, and open discussion content. |
| docs/ | General documentation, such as usage guides, FAQs, operation manuals, etc. |
| proposals/ | Requirement analysis, future planning, and proposal documents. |
| principles/ | Technical principle analysis, underlying mechanism explanation documents. |
| references/ SIG references to processes, specifications, and other documents in this community or other communities. |
Participating and Contributing Code
Please refer to Contribution Guide