nvidia docker commands

Useful command

ddokkddokk 2017. 8. 16. 15:52

# DOCKER 사용법

---------------------------------

>> putty로 서버에 접속

IP: 129.254.121.81, PORT: 22

---------------------------------

>> super user login

$ sudo su

---------------------------------

>> docker image 확인

$ nvidia-docker images

---------------------------------

>> container 생성& 폴더 마운트

$ nvidia-docker run -it -d -p (OS포트):8888 --name container_name -v /data:/data docker_image_id

ex)

nvidia-docker run -it -d -p 18888:8888 --name torcs_v0 -v /data:/data e86

nvidia-docker run -it -d -p 28888:8888 --name torcs_v1 -v /data:/data e86

nvidia-docker run -it -d -p 38888:8888 --name torcs_v2 -v /data:/data e86

nvidia-docker run -it -d -p 48888:8888 --name torcs_v3 -v /data:/data e86

---------------------------------

>> 기 생성된 container 확인

$ nvidia-docker ps -a

---------------------------------

>> contrainer 실행

$ nvidia-docker exec -it container_id_or_name /bin/bash

ex) nvidia-docker exec -it torcs_v0 /bin/bash

---------------------------------

>> container 실행 중지

$nvidia-docker stop container_id_or_name

---------------------------------

>> container 삭제

$nvidia-docker rm container_id_or_name

# jupyter사용법 (/data/notebook 폴더에 소스파일 저장)

접속주소 : http://129.254.121.81:8888/