3 Pull requests merged by 2 people

• Switch to context package in go stdlib

  #1114 merged Jan 10, 2025

• Raise a warning instead of an error if GPU mode labeler fails

  #1106 merged Jan 10, 2025

• Add ada-lovelace architecture label for compute capability 8.9

  #1090 merged Jan 9, 2025

7 Pull requests opened by 3 people

• Expose device UUIDs to node label

  #1116 opened Jan 9, 2025

• Bump github.com/NVIDIA/go-nvml from 0.12.4-0 to 0.12.4-1

  #1117 opened Jan 12, 2025

• Bump google.golang.org/grpc from 1.65.0 to 1.69.4

  #1120 opened Jan 13, 2025

• Bump sigs.k8s.io/node-feature-discovery/api/nfd from 0.16.6 to 0.17.1

  #1122 opened Jan 14, 2025

• Bump sigs.k8s.io/node-feature-discovery/api/nfd from 0.16.6 to 0.17.1 in /tests

  #1123 opened Jan 14, 2025

• Bump the k8sio group across 2 directories with 5 updates

  #1124 opened Jan 16, 2025

• do not need to set default runtime nvidia in docker

  #1125 opened Jan 16, 2025

1 Issue closed by 1 person

• Containers that use cuda images in k8s do not have gpu resources, but the process id can be seen using nvidia-smi

  #954 closed Jan 12, 2025

4 Issues opened by 4 people

• nvidia-device-plugin fails to recover after apiserver is inaccessible for a moment

  #1126 opened Jan 16, 2025

• Why disable resource naming config?

  #1121 opened Jan 13, 2025

• Feature Request: MPS memory control / oversubscription

  #1119 opened Jan 13, 2025

• MPS use error: Failed to allocate device vector A (error code all CUDA-capable devices are busy or unavailable)!

  #1118 opened Jan 13, 2025

7 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Add gpu uuids to node labels

  #1015 commented on Jan 9, 2025 • 0 new comments

• GPU resources are not recovered even XID error is resolved

  #1014 commented on Jan 10, 2025 • 0 new comments

• Enable hostNetwork = true

  #962 commented on Jan 10, 2025 • 0 new comments

• Support automatic discovery of MIG devices

  #992 commented on Jan 14, 2025 • 0 new comments

• k8s pod ,After running for a while, the GPU cannot be found in the pod. Failed to initialize NVML: Unknown Error

  #981 commented on Jan 15, 2025 • 0 new comments

• If all devices are pre-volta, skip setting set_default_device_pinned_mem_limit and set_default_active_thread_percentage.

  #971 commented on Jan 14, 2025 • 0 new comments

• Bump the k8sio group across 1 directory with 5 updates

  #1100 commented on Jan 12, 2025 • 0 new comments