fgci-org/ansible-role-cuda

Issues

• run cuda_init.sh before ansible-pull-script.sh

  #3 opened 8 years ago by martbhell
  0

• Specifying cuda version doesn't work

  #29 opened 5 years ago by tgamauf
  4

• Initialize GPUs task fails on VM with passthrough GPU

  #23 opened 5 years ago by verdurin
  6

• Improve score on ansible galaxy

  #26 opened 5 years ago by martbhell
  6

• Galaxy latest version is 1.5.4

  #24 opened 5 years ago by uda
  2

• Persistence mode is deprecated

  #19 opened 5 years ago by jabl
  1

• Need to support nvidia legacy driver

  #16 opened 6 years ago by jabl
  3

• Add self-healing for when driver/kernel don't match

  #14 opened 7 years ago by martbhell
  0

• deprecation warning when running role

  #13 opened 7 years ago by victorklos
  2

• testing issue: ubuntu 1604 and avahi and lxc

  #11 opened 8 years ago by martbhell
  6

• Still problems with new CUDA repo key

  #6 opened 8 years ago by jabl
  7

• Remove the lineinfile task to rc.local to run cuda_init.sh on boot

  #5 opened 8 years ago by martbhell
  1

• skip wait_for in reboot handler ( for ansible-pull )

  #4 opened 8 years ago by martbhell
  4

• load /dev/nvidia0 before slurm role runs

  #2 opened 8 years ago by A1ve5
  1

• Reboot node after CUDA install fails when using a bastion host

  #1 opened 9 years ago by martbhell
  4