1. welcome
  2. Hello World
  3. Open Source
  4. 1. vllm
  5. 2. sglang
  6. 3. pytorch
    ❱
    1. 3.1. torchrun
    2. 3.2. tensor
    3. 3.3. autograd
    4. 3.4. operator
    5. 3.5. profiler
    6. 3.6. hook
    7. 3.7. elastic
    8. 3.8. patch
    9. 3.9. misc
  7. 4. paddlepaddle
    ❱
    1. 4.1. ps
    2. 4.2. framework
    3. 4.3. cinn
    4. 4.4. dataloader
  8. 5. horovod
    ❱
    1. 5.1. run
    2. 5.2. workflow
    3. 5.3. object
    4. 5.4. develop
    5. 5.5. pytorch
    6. 5.6. tensorflow
    7. 5.7. elastic
  9. 6. ray
    ❱
    1. 6.1. overview
    2. 6.2. gcs
    3. 6.3. raylet
    4. 6.4. api
    5. 6.5. survey
  10. 7. llama
  11. 8. nccl
  12. 9. megatron
  13. 10. deepspeed
  14. 11. nanochat
  15. Survey
  16. 12. survey
    ❱
    1. 12.1. pollux
    2. 12.2. adasum
    3. 12.3. adaptation_learning
    4. 12.4. gradient_descent
    5. 12.5. auto_parallel
    6. 12.6. scheduling
    7. 12.7. gradient_compression
      ❱
      1. 12.7.1. dgc
      2. 12.7.2. csc
    8. 12.8. flash attention
    9. 12.9. LoRA
  17. 13. models
    ❱
    1. 13.1. llm
    2. 13.2. falcon
    3. 13.3. llama
    4. 13.4. peft
    5. 13.5. transformer
    6. 13.6. models
  18. Programming
  19. 14. python
    ❱
    1. 14.1. concurrent execution
    2. 14.2. multiprocessing
    3. 14.3. decorator
  20. 15. golang
    ❱
    1. 15.1. golang error
  21. 16. cplusplus
    ❱
    1. 16.1. enable_shared_from_this
  22. Mathematics
  23. 17. mathematics
    ❱
    1. 17.1. basic
    2. 17.2. entropy
    3. 17.3. newton
    4. 17.4. regression
    5. 17.5. conjugate descent
    6. 17.6. gradient descent
    7. 17.7. pca
    8. 17.8. support vector
    9. 17.9. differentiation
    10. 17.10. fourier
    11. 17.11. kmeans
  24. 18. wavelets
    ❱
    1. 18.1. plan
    2. 18.2. preliminary
    3. 18.3. haar wavelet
    4. 18.4. fourier analysis
    5. 18.5. uncertainty principle
    6. 18.6. multiresolution
  25. Learning Deep
  26. 19. kubernetes
    ❱
    1. 19.1. concepts
    2. 19.2. scheduler
    3. 19.3. operator
    4. 19.4. device plugin
    5. 19.5. docker
    6. 19.6. install
    7. 19.7. api-service
    8. 19.8. controller
  27. 20. cuda
    ❱
    1. 20.1. 101
  28. 21. todo
    ❱
    1. 21.1. gloo
    2. 21.2. mpi
    3. 21.3. jax
    4. 21.4. tvm
    5. 21.5. llm
  29. 22. notes
    ❱
    1. 22.1. influence and persuasion
    2. 22.2. freynman technique
    3. 22.3. wavelet signal processing
  30. 23. tips
    ❱
    1. 23.1. ip_local_port_range
  31. 24. infrastructure
    ❱
    1. 24.1. pki
    2. 24.2. linux cache
  32. 25. projects
    ❱
    1. 25.1. copilot
    2. 25.2. library
    3. 25.3. RAG
  33. 26. chronicles
    ❱
    1. 26.1. feb 2024

Aller au boulot

scheduler

Summary and Examples

比较简单的 demo,体现基本框架

GitHub - Huang-Wei/sample-scheduler-extender: a sample to showcase how to create a k8s scheduler extender

稍微复杂一些,完整实现,extender 部署

GitHub - everpeace/k8s-scheduler-extender-example: An example of kubernetes scheduler extender

aliyun gpu,部署替换 kube-scheduler

GitHub - AliyunContainerService/gpushare-scheduler-extender: GPU Sharing Scheduler for Kubernetes Cluster

基于 kubernetes framework

GitHub - everpeace/kube-throttler: throttling your pods in kubernetes cluster.

Reference

  • design-proposals-archive/scheduler_extender.md at main · kubernetes/design-proposals-archive · GitHub

  • Scheduling Framework | Kubernetes

  • enhancements/README.md at master · kubernetes/enhancements · GitHub