본문 바로가기
Nvidia/TEST

nvidia-smi drain 명령어

by ccclog 2024. 9. 1.
반응형

명령어에 대한 설명을 온라인에서 찾기 어려워서 작성해둡니다.

 

특정 GPU의 드라이버상에서 인식을 Disable할 때 사용했습니다.

 

#nvidia-smi

root@user:~# nvidia-smi

 

 

#nvidia-smi drain

 

root@user:~# nvidia-smi drain

 

    drain -- Display drain state information about the system as well as remove and discover devices.

 

    Usage: nvidia-smi drain [options]

 

    Options include:

    [-p | --pciid]: GPU PCI ID in the format XXXX:YY.Z.a

                   where XXXX = domain

                         YY   = bus

                         Z    = device                         a    = function

    [-m | --modify]: Modify the drain state of a GPU specified by -p.

   Persistence mode must be disabled to perform this function.

                   0 = not draining

                   1 = draining

    [-q | --query]: Query the drain state of a GPU specified by -p.

    [-r | --remove]: If possible, remove the GPU specified by -p.

                   The GPU must have already been put in a drain state using -m and

                   persistence mode for this GPU must be disabled.

 

    [-d | --discover]: Discover all GPUs on the system that had previously been removed.

    [-h | --help]: Display help information

 

 

# drain 적용.

root@user:~# nvidia-smi drain -p 0000:18:00.0 -m 1

Successfully set GPU 00000000:18:00.0 drain state to: draining.

 

 

#결과1

root@user:~# nvidia-smi

No devices were found

 

#drain 해제

root@user:~# nvidia-smi drain -p 0000:18:00.0 -m 0

Successfully set GPU 00000000:18:00.0 drain state to: not draining.

 

#결과2

root@user:~# nvidia-smi

 

반응형

'Nvidia > TEST' 카테고리의 다른 글

DCGMI repository 등록 및 설치/실행(Ubuntu 22.04.3)  (0) 2023.10.18
NVLINK 상태 확인  (0) 2023.09.23
GPU-BURN  (0) 2022.11.14
Nvidia-bug-report.sh  (0) 2022.11.03