docker筆記.md

tensorflow GPU docker

nvidia docker照著nvidia的github裝好像會噴錯誤訊息

可以嘗試用這個看看 某部落格

1. First, you need to install Docker:
   • curl -fsSL get.docker.com -o get-docker.sh
   • sh get-docker.sh
   • Use docker without sudo:sudo usermod -aG docker $USER
   • Now, have the user logout then login again.
2. Second, install tensorflow docker image:
   • Follow this README, you also need to install nvidia-docker which has details in README too. https://github.com/tensorflow/tensorflow/tree/master/tensorflow/tools/docker
3. Create your own container:
   • Python3 + GPU:nvidia-docker run -itd --name <name> -p <port>:8888 tensorflow/tensorflow:nightly-gpu-py3 /run_jupyter.sh --allow-root --ip=0.0.0.0
   • Python2 + GPU:nvidia-docker run -itd --name <name> -p <port>:8888 tensorflow/tensorflow:nightly-gpu /run_jupyter.sh --allow-root --ip=0.0.0.0

銘毅的ssh ubuntu dockerfile

Dockerfile

FROM ubuntu:14.04
MAINTAINER davidtnfsh


RUN sudo apt-get update && \
        sudo apt-get install -y openssh-server \
                                            openjdk-7-jdk \
                                            wget \
                                            vim \


# SSH�K�˙C����֧Ԯ
RUN sudo locale-gen zh_TW zh_TW.UTF-8 
RUN echo "LC_CTYPE=zh_TW.UTF-8" | sudo tee -a /etc/environment

CMD ["sh","-c" ,"/etc/init.d/ssh start; bash"]

create docker sh

#!/bin/bash
read -p "Create UserNmae  : " USER_NAME

read -p "Setting Password  [ default : udiclab@720 ] : " USER_PWD
if [ ${#USER_PWD} == 0 ] ; then
	USER_PWD="udiclab@720"
fi

read -p "Setting SSH's Port  : " USER_PORT

if [ ${#USER_NAME} == 0 -o ${#USER_PORT} == 0 ] ; then
	echo "Please at least enter the user name and the port number"
	exit 0
fi

echo -e "\nYou Setting:"
echo "User: ${USER_NAME}"
echo "Password: ${USER_PWD}"
echo "Port: ${USER_PORT}"

sudo docker run -itd \
                -p $USER_PORT:22 \
                --name $USER_NAME \
                --hostname $USER_NAME \
                udic/ubuntu:1.0 &> /dev/null

# sudo docker exec -ti $USER_NAME sudo sh -c "sudo apt-get update && sudo apt-get upgrade && sudo apt-get install -y openssh-server"

sudo docker exec -ti $USER_NAME sudo sh -c "sudo useradd -m $USER_NAME -s /bin/bash;echo \"${USER_NAME}:${USER_PWD}\" | chpasswd;sudo adduser $USER_NAME sudo" &> /dev/null

# sudo docker exec -ti $USER_NAME sudo sh -c "sudo /etc/init.d/ssh start"

# 終端機中文支援
# sudo docker exec -ti $USER_NAME sudo sh -c "sudo locale-gen zh_TW zh_TW.UTF-8;echo \"LC_CTYPE=zh_TW.UTF-8\" | sudo tee -a /etc/environment" &> /dev/null

刪除docker指令

1. 刪除同一個image創出來的container指令： docker ps -a | awk '{ print $1,$2 }' | grep centos:7 | awk '{print $1 }' | xargs -I {} docker rm {}
2. 刪除特定名稱的images(通常是建images失敗都會變成none的名稱)：docker images | awk '{ print $1,$2,$3 }' | grep "<none>" | awk '{print $3}' | xargs -I {} docker rmi {}
3. 嘗試建立image的時候，前幾次失敗的都會變成hash值，一口氣刪除的指令：

docker rmi `docker images --filter dangling=true -q` #Removes all those images.

4. 將container容量排序： docker ps -a --format '{{.Size}}\t{{.Names}}' | sed 's/ //' | sort -h -r
5. 將image容量做排序： docker images --format '{{.Size}}\t{{.Repository}}' | sed 's/ //' | sort -h -r
6. remove dangling volume:

   docker volume rm `docker volume ls -q -f dangling=true`