ROCm/k8s-device-plugin

Issues

• Device permissions set by the device-plugin cause unexpected access() syscall responses, ending up in Pytorch failures

  #65 opened 5 months ago by elukey
  8

• Runtime Error with AMD GPU Helm Chart Installation in Kubernetes

  #48 opened 10 months ago by maarten-blokker
  1

• [Feature]: Sharing without restrictions

  #66 opened 5 months ago by sfxworks
  0

• [Documentation]: Introduction link broken

  #63 opened 6 months ago by chipzoller
  0

• [Issue]: namespace configuration in Helm chart

  #60 opened 6 months ago by rptaylor
  0

• question about privileged mode for labeller

  #59 opened 6 months ago by rptaylor
  4

• [Issue]: need a way to overwrite node_selector labels

  #61 opened 6 months ago by rptaylor
  0

• Permissions for the /dev/{kfd,dri/renderXXXX} devices in containers

  #39 opened 2 years ago by elukey
  1

• [Documentation]: clarification on kernel/driver installation

  #57 opened 7 months ago by rptaylor
  1

• [Documentation]: MI200 support

  #56 opened 7 months ago by rptaylor
  1

• [Feature]: Support detection, allocation and resetting of GPU partitions in CDNA cards

  #54 opened 8 months ago by lohbe
  0

• Does this device plugin support kubevirt about GPU?

  #25 opened 3 years ago by larryliuqing
  1

• Is AMD Radeon Vega 8 supported?

  #42 opened 10 months ago by dmfrey
  18

• [Issue]: benchmark example is broken

  #52 opened 9 months ago by andy108369
  1

• [Issue]: ArtifactHUB helm install points to non-existant chart repo

  #51 opened 9 months ago by jkoelker
  1

• [Issue]: Helm Chart unavailable

  #49 opened 9 months ago by LucaDev
  2

• 404 link in README.md

  #43 opened 10 months ago by hy-tomas-terala
  0

• GPU isolation options

  #45 opened 10 months ago by andy108369
  0

• Libraries/binaries mounted in the container (analogous to NVIDIA_DRIVER_CAPABILITIES)

  #44 opened 10 months ago by andy108369
  0

• Registration failing on K3s

  #31 opened 2 years ago by Scrumplex
  12

• EKS Support

  #40 opened a year ago by aldmbmtl
  0

• Node detected and labelled but has no capacity?

  #33 opened 2 years ago by agilob
  8

• Is it possible to restrict vram/cu per pod?

  #9 opened 4 years ago by hyc3z
  2

• add podLabels and podAnnotations to labeller's and device plugin's pods

  #37 opened 2 years ago by boniek83
  0

• device plugin daemonset tries to load itself on non-amd64 nodes

  #34 opened 2 years ago by agilob
  1

• Dead chart repo. Please fix

  #32 opened 2 years ago by MrAmbiG
  1

• Issues in helm chart

  #19 opened 3 years ago by boniek83
  1

• Helm Incompitable Version with AWS EKS Cluster

  #24 opened 2 years ago by utkudarilmaz
  0

• Prometheus metrics of gpu resources

  #15 opened 3 years ago by boniek83
  1

• Talos Vega 11 Fail to glob topology properties

  #23 opened 3 years ago by anthr76
  3

• Not Working On Ubuntu 20.04 Kernel 5.8 HWE

  #21 opened 3 years ago by sstubbs
  3

• would all pods using the same GPU are assigned to the same NUMA?

  #20 opened 3 years ago by mrpre
  1

• Create helm chart repository

  #18 opened 3 years ago by boniek83
  1

• wrong

  #14 opened 4 years ago by ccnankai
  0

• Is amdgpu pro required?

  #11 opened 4 years ago by pschmitt
  1

• Can't find any GPU

  #12 opened 4 years ago by qiyueyuanwei
  3

• [Q] Pod can specify GPU device-id?

  #10 opened 4 years ago by sakaia
  1

• Deprecated k8s apiversion in yaml

  #7 opened 5 years ago by abuccts
  1

• New release compatible with k8s 1.10?

  #1 opened 6 years ago by eriklindahl
  1