前言一、分区
1.创建分区2.格式化分区 二、lxd
1.安装2.初始化3.容器创建4.配置ssh5.升级ubuntu6.添加显卡7.安装驱动8.安装图形化界面9.制作镜像10.共享存储卷 总结
前言
由于之前的服务器被我搞坏了所以要重新开始配置服务器lxc管理
一、分区 1.创建分区
目的：划分出3T的分区来挂载lxc
查看磁盘
sudo fdisk -l
Disk /dev/sda: 3.7 TiB, 4000787030016 bytes, 7814037168 sectors
这一块3.7T的就是要进行分区操作的盘，然后由于默认是mbr格式，无法通过fdisk划分2T以上的分区，因此用parted工具将mbr转化为gpt
sudo parted /dev/sda #进入parted mklabel gpt #将磁盘设置为gpt格式 mkpart logical 0 -1 #将磁盘所有的容量设置为GPT格式，这里会格式化丢失原先文件，请慎重 print #查看分区结果
这个时候应该是默认进行分了一个/dev/sda1这个分区，这个分区包含了整个磁盘大小，不是我们想要的3T，因此要删除重新配置
rm 1 #删除编号为1的分区 unit GB #改变默认单位GB mkpart #进入创立分区交互指令 Partition name? []? pv1 File system type? [ext2]? zfs Start? 0 End? 3072GB print #分区完成后打印出来看一下结果 Model: ATA HGST HUS726T4TAL (scsi) Disk /dev/sda: 4001GB Sector size (logical/physical): 512B/4096B Partition Table: gpt Disk Flags: Number Start End Size File system Name Flags 1 1049kB 3072GB 3072GB zfs pv1
在parted中完成了分区，然后通过fdisk验证一下
sudo fdisk -l
Device Start End Sectors Size Type /dev/sda1 2048 6000001023 5999998976 2.8T Linux filesystem
fdisk中也可以看到对应的分区结果，至此完成了分区。
2.格式化分区将刚刚分出来的sda1格式化为ext4的格式
sudo mkfs -t ext4 /dev/sda1
mke2fs 1.44.1 (24-Mar-2018)
Creating filesystem with 749999872 4k blocks and 187506688 inodes
Filesystem UUID: a66ff254-e95f-4981-921e-a090e7aff4a1
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
102400000, 214990848, 512000000, 550731776, 644972544
Allocating group tables: done
Writing inode tables: done
Creating journal (262144 blocks): done
Writing superblocks and filesystem accounting information: done
c
查看硬盘/dev/sda1 对应的UUID
sudo blkid
注意: 唯一的sda1的UUID号。再事先准备好一个地方来做挂载点，比如我这里是/data1然后再用命令打开配置文件：
sudo vim /etc/fstab
然后在文件末尾添加
UUID=a66ff254-e95f-4981-921e-a090e7aff4a1 /data1 ext4 defaults 0 0
重启
df
Filesystem 1K-blocks Used Available Use% Mounted on udev 65774384 0 65774384 0% /dev tmpfs 13166396 10880 13155516 1% /run /dev/sdb2 921923300 46802332 828220132 6% / tmpfs 65831968 0 65831968 0% /dev/shm tmpfs 5120 4 5116 1% /run/lock tmpfs 65831968 0 65831968 0% /sys/fs/cgroup /dev/loop0 113536 113536 0 100% /snap/core/12725 tmpfs 13166392 0 13166392 0% /run/user/1000 /dev/sda1 2951859536 90140 2801753040 1% /data1
可以看到/dev/sda1已经成功挂载到了/data1下。证明了开机自动挂载完成。最后再测验一下卸载和挂载功能
sudo umount -l /dev/sda1 df
Filesystem 1K-blocks Used Available Use% Mounted on udev 65774384 0 65774384 0% /dev tmpfs 13166396 10880 13155516 1% /run /dev/sdb2 921923300 46802332 828220132 6% / tmpfs 65831968 0 65831968 0% /dev/shm tmpfs 5120 4 5116 1% /run/lock tmpfs 65831968 0 65831968 0% /sys/fs/cgroup /dev/loop0 113536 113536 0 100% /snap/core/12725 tmpfs 13166392 0 13166392 0% /run/user/1000
没有找到/dev/sda1，卸载成功
sudo mount /dev/sda1 /data1 df
Filesystem 1K-blocks Used Available Use% Mounted on udev 65774384 0 65774384 0% /dev tmpfs 13166396 10880 13155516 1% /run /dev/sdb2 921923300 46802332 828220132 6% / tmpfs 65831968 0 65831968 0% /dev/shm tmpfs 5120 4 5116 1% /run/lock tmpfs 65831968 0 65831968 0% /sys/fs/cgroup /dev/loop0 113536 113536 0 100% /snap/core/12725 tmpfs 13166392 0 13166392 0% /run/user/1000 /dev/sda1 2951859536 90140 2801753040 1% /data1
又重新挂载上了，至此分区工作完成。
二、lxd 1.安装lxd是lxc的升级版，安装包名为lxd，命令仍然沿用lxc。
lxd的安装方式有两种，apt和snap，下面列举了两者的差异。
apt比较熟悉，刚开始想通过apt安装，但是安装过程中一直提示连接不上snap导致无法安装。由此推断apt也是封装snap命令进行安装lxd，所以直接采用snap进行安装，并且选用推荐的4.x版本。
sudo snap install lxd
安装完成后可以通过snap version看到对应的版本信息
ssme@ssme-server:~$ snap version snap 2.54.3.2 snapd 2.54.3.2 series 16 ubuntu 20.04 kernel 5.4.0-100-generic
当然过程中不一定是一帆风顺的，我遇到过两个问题。
第一次提示我
connect: connection refused
这是网络连接问题，查看网络是否可用，我通过设置代理解决了
第二此提示我
read: connection reset by peer
字面意思是连接被同行重置，理解一下就是网络配置被修改了。我因为修改过代理，并且是通过export http设置，vim ~/.bashrc以及snap自带的vim /etc/environment 等三种方式修改了代理，虽然代理是同一个ip，但是经过无数次的失败后，最后尝试将之前的修改的代理只保留一个。
export proxy可以通过重启或者export http_proxy="",设置为空来清除~/.bashrc 保留/etc/environment 删除proxy配置信息
然后重新尝试安装，发现安装完成，坑爹啊！就算是完成了，我依然感觉这是个bug。
2.初始化首先把之前挂载的分区拿下来，这里不需要挂，lxc有一套自己的挂载方式
umount /dev/sda1
在块设备 /dev/sda1 上创建一个ZFS存储池
sudo lxc storage create zfs-pool zfs source=/dev/sda1
创建成功后查看一下,因为没有配置占用空间，所以默认占用全部大约2.7T
ssme@ssme-server:~$ sudo lxc storage info zfs-pool info: description: "" driver: zfs name: zfs-pool space used: 1.56MiB total space: 2.69TiB
lxd初始化，第二轮互动询问是否要创建资源池，这里我们通过指定分区创建资源池了，所以不需要默认创建（如果默认创建的话就会挂载到别的分区去）
ssme@ssme-server:~$ sudo lxd init
Would you like to use LXD clustering? (yes/no) [default=no]:
Do you want to configure a new storage pool? (yes/no) [default=yes]: no
Would you like to connect to a MAAS server? (yes/no) [default=no]:
Would you like to create a new local network bridge? (yes/no) [default=yes]:
What should the new bridge be called? [default=lxdbr0]:
The requested network bridge "lxdbr0" already exists. Please choose another name.
What should the new bridge be called? [default=lxdbr0]: lxdbr0
The requested network bridge "lxdbr0" already exists. Please choose another name.
What should the new bridge be called? [default=lxdbr0]: lxdbb1
What IPv4 address should be used? (CIDR subnet notation, “auto” or “none”) [default=auto]:
What IPv6 address should be used? (CIDR subnet notation, “auto” or “none”) [default=auto]:
Would you like the LXD server to be available over the network? (yes/no) [default=no]:
Would you like stale cached images to be updated automatically? (yes/no) [default=yes]:
Would you like a YAML "lxd init" preseed to be printed? (yes/no) [default=no]: yes
config: {}
networks:
- config:
ipv4.address: auto
ipv6.address: auto
description: ""
name: lxdbb1
type: ""
project: default
storage_pools: []
profiles:
- config: {}
description: ""
devices:
eth0:
name: eth0
network: lxdbb1
type: nic
name: default
projects: []
cluster: null
再次配置，修改网桥（由于我上面拼写错了，所以这里顺带修改了）
sudo lxc profile edit default3.容器创建
目的：采用清华源中的ubuntu镜像创建容器，并通过ssh连接
添加清华镜像源地址
sudo lxc remote add tuna-images https://mirrors.tuna.tsinghua.edu.cn/lxc-images/ --protocol=simplestreams --public
查看该仓库中可用镜像
sudo lxc image list tuna-images:
使用清华源中的ubuntu镜像创建一个叫alpha的容器，这里我报错，网上资料显示应该是配置文件出了问题
ssme@ssme-server:~$ sudo lxc launch tuna-images:ubuntu/18.04 alpha Creating alpha Error: Failed instance creation: Failed creating instance record: Failed initialising instance: Invalid devices: Failed detecting root disk device: No root device could be found
按照要求修改配置文件，这里我添加了root一栏，下面设置了根目录位置以及最大空间。
sudo lxc profile edit default
然后我再次创建容器alpha，创建成功
ssme@ssme-server:~$ sudo lxc launch tuna-images:ubuntu/18.04 alpha Creating alpha Starting alpha
进入容器
sudo lxc exec test bash
我们登录的是root用户，在这个容器中已经存在了一个叫ubuntu的用户
修改密码
passwd root passwd ubuntu
安装ssh
apt install ssh
通过ssh连接容器，查看容器与宿主机的ip，因为我们没有设置桥接网卡，不能从外部电脑访问容器（不能ping通容器的ip），因此我们采用端口监听的方式来访问我们的容器。
退出容器
exit
在宿主机查看容器
ssme@ssme-server:~$ sudo lxc list [sudo] password for ssme: +-------+---------+----------------------+------+-----------+-----------+ | NAME | STATE | IPV4 | IPV6 | TYPE | SNAPSHOTS | +-------+---------+----------------------+------+-----------+-----------+ | alpha | RUNNING | 10.246.13.238 (eth0) | | ConTAINER | 0 | +-------+---------+----------------------+------+-----------+-----------+
图中可知容器的ip地址为10.246.13.238
查看宿主机ip地址
ip addr
可知宿主机ip为10.83.17.189
端口转发
ssme@ssme-server:~$ sudo lxc config device add alpha proxy0 proxy listen=tcp:10.83.17.189:60601 connect=tcp:10.246.13.238:22 bind=host [sudo] password for ssme: Device proxy0 added to alpha
60601是我们定的端口号，通过宿主机的60601端口号监听容器中22端口号（SSH默认端口号）完成端口绑定，这样我们就可以通过ssh直接连接容器，同时可以查看对应的容器配置文件，发现已经将端口转发配置固定下来了
sudo lxc config edit alpha
architecture: x86_64
config:
image.architecture: amd64
image.description: Ubuntu bionic amd64 (20220228_07:42)
image.os: Ubuntu
image.release: bionic
image.serial: "20220228_07:42"
image.type: squashfs
image.variant: default
volatile.base_image: 04d0148a87275d123d7d0972047d1216330a39513796caff1f1827943d3e7154
volatile.eth0.host_name: veth9948d26c
volatile.eth0.hwaddr: 00:16:3e:95:6e:3f
volatile.idmap.base: "0"
volatile.idmap.current: '[{"Isuid":true,"Isgid":false,"Hostid":1000000,"Nsid":0,"Maprange":1000000000},{"Isuid":false,"Isgid":true,"Hostid":1000000,"Nsid":0,"Maprange":1000000000}]'
volatile.idmap.next: '[{"Isuid":true,"Isgid":false,"Hostid":1000000,"Nsid":0,"Maprange":1000000000},{"Isuid":false,"Isgid":true,"Hostid":1000000,"Nsid":0,"Maprange":1000000000}]'
volatile.last_state.idmap: '[{"Isuid":true,"Isgid":false,"Hostid":1000000,"Nsid":0,"Maprange":1000000000},{"Isuid":false,"Isgid":true,"Hostid":1000000,"Nsid":0,"Maprange":1000000000}]'
volatile.last_state.power: RUNNING
volatile.uuid: 1333a63f-0370-4a5a-94a1-bc62ecde9f3c
devices:
proxy0:
bind: host
connect: tcp:10.246.13.238:22
listen: tcp:10.83.17.189:60601
type: proxy
ephemeral: false
profiles:
- default
stateful: false
对应字段解释请看下图
后面还要配置显卡，磁盘大小，都会相应的写入配置文件中
使用ssh连接容器成功
ssme@ssme-server:~$ sudo ssh ubuntu@10.83.17.189 -p 606015.升级ubuntu
更换apt源
sudo mv /etc/apt/sources.list /etc/apt/sources.list.bak sudo vim /etc/apt/sources.list
网易源 bionic版本
deb http://mirrors.163.com/ubuntu/ bionic main restricted universe multiverse deb-src http://mirrors.163.com/ubuntu/ bionic main restricted universe multiverse deb http://mirrors.163.com/ubuntu/ bionic-security main restricted universe multiverse deb-src http://mirrors.163.com/ubuntu/ bionic-security main restricted universe multiverse deb http://mirrors.163.com/ubuntu/ bionic-updates main restricted universe multiverse deb-src http://mirrors.163.com/ubuntu/ bionic-updates main restricted universe multiverse deb http://mirrors.163.com/ubuntu/ bionic-proposed main restricted universe multiverse deb-src http://mirrors.163.com/ubuntu/ bionic-proposed main restricted universe multiverse deb http://mirrors.163.com/ubuntu/ bionic-backports main restricted universe multiverse deb-src http://mirrors.163.com/ubuntu/ bionic-backports main restricted universe multiverse
然后更新源，养成好习惯
sudo apt update
查看一下内核版本和发行版本
ubuntu@alpha:~$ cat /proc/version Linux version 5.4.0-100-generic (buildd@lcy02-amd64-002) (gcc version 9.3.0 (Ubuntu 9.3.0-17ubuntu1~20.04)) #113-Ubuntu SMP Thu Feb 3 18:43:29 UTC 2022 ubuntu@alpha:~$ lsb_release -a No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 18.04.6 LTS Release: 18.04 Codename: bionic
发现内核版本和发行版本对应不上，担心以后会出现问题，所以直接升级成ubuntu20，操作比较简单，过程就是一直默认往下进行
apt update apt upgrade apt dist-upgrade do-release-upgrade
更新成功以后重启再查看发行版版本
ubuntu@alpha:~$ lsb_release -a No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 20.04.4 LTS Release: 20.04 Codename: focal
成功升级到ubuntu20focal版本，再换源，养成好习惯
# 默认注释了源码镜像以提高 apt update 速度，如有需要可自行取消注释 deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-updates main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-updates main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-backports main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-backports main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-security main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-security main restricted universe multiverse # 预发布软件源，不建议启用 # deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-proposed main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ focal-proposed main restricted universe multiverse
采用清华源，注意版本ubuntu20 focal。换源步骤同上。
6.添加显卡为容器添加所有GPU:
lxc config device add alpha gpu gpu
添加指定GPU：
先查看gpu阵列
ssme@ssme-server:~$ nvidia-smi Fri Mar 4 09:41:17 2022 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 470.42.01 Driver Version: 470.42.01 CUDA Version: 11.4 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA GeForce ... On | 00000000:18:00.0 Off | N/A | | 16% 29C P8 3W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 1 NVIDIA GeForce ... On | 00000000:3B:00.0 Off | N/A | | 16% 29C P8 7W / 250W | 18MiB / 11018MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 2 NVIDIA GeForce ... On | 00000000:86:00.0 Off | N/A | | 16% 29C P8 9W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 3 NVIDIA GeForce ... On | 00000000:AF:00.0 Off | N/A | | 16% 29C P8 15W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | 0 N/A N/A 1561 G /usr/lib/xorg/Xorg 4MiB | | 1 N/A N/A 1561 G /usr/lib/xorg/Xorg 9MiB | | 1 N/A N/A 1646 G /usr/bin/gnome-shell 6MiB | | 2 N/A N/A 1561 G /usr/lib/xorg/Xorg 4MiB | | 3 N/A N/A 1561 G /usr/lib/xorg/Xorg 4MiB | +-----------------------------------------------------------------------------+
把第一张卡分给alpha
lxc config device add alpha gpu0 gpu id=0
或者，这里pci中第一段8个0要换成4个0，可能是lxc只能读取四位，全部复制会有bug
lxc config device add jellyfish gpu gpu pci=0000:18:00.07.安装驱动
添加好显卡后，就相当于我们给容器安装了显卡，我们回到容器，然后安装显卡驱动与宿主机的显卡版本必须一致，安装方法参照我的另一篇博客ubuntu下安装nividia驱动
ubuntu上安装非常简单
ubuntu@alpha:~$ nvidia-smi Wed Mar 2 05:42:54 2022 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 470.42.01 Driver Version: 470.42.01 CUDA Version: 11.4 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA GeForce ... On | 00000000:18:00.0 Off | N/A | | 20% 37C P8 4W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 1 NVIDIA GeForce ... On | 00000000:3B:00.0 Off | N/A | | 19% 38C P8 9W / 250W | 18MiB / 11018MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 2 NVIDIA GeForce ... On | 00000000:86:00.0 Off | N/A | | 19% 37C P8 8W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ | 3 NVIDIA GeForce ... On | 00000000:AF:00.0 Off | N/A | | 19% 37C P8 15W / 250W | 5MiB / 11019MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================|
再验证nvcc发现出错
ubuntu@alpha:~$ nvcc -v -bash: nvcc: command not found
应该是没有把cuda写到环境变量中，打开bashrc，添加nvcc的文件夹位置
ubuntu@alpha:/usr/local/cuda/bin$ sudo vim ~/.bashrc
在结尾处加入 export PATH=$PATH:/usr/local/cuda/bin
再激活一下，nvcc就配置好了。
ubuntu@alpha:/usr/local/cuda/bin$ source ~/.bashrc ubuntu@alpha:/usr/local/cuda/bin$ nvcc -V nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2021 NVIDIA Corporation Built on Wed_Jun__2_19:15:15_PDT_2021 Cuda compilation tools, release 11.4, V11.4.48 Build cuda_11.4.r11.4/compiler.30033411_08.安装图形化界面
刷新源
sudo apt update
安装无推荐软件的ubuntu桌面(默认安装gnome，完整安装会有很多无关的软件)
sudo apt install --no-install-recommends ubuntu-desktop
安装远程连接使用安装脚本（安装git后下载我们之后需要用的东西）
sudo apt install git git clone https://github.com/shenuiuin/LXD_GPU_SERVER
打开文件夹，赋予脚本可执行权限
cd LXD_GPU_SERVER/ sudo chmod a+x xrdp-installer-1.2.3.sh
脚本会下载一些文件，需要有Downloads文件夹
mkdir -p ~/Downloads
安装脚本
./xrdp-installer-1.2.3.sh -c -l -s
安装完成
退出容器，添加端口绑定，命令跟ssh差不多，在服务器上远程连接测试
ssme@ssme-server:~$ lxc config device add alpha proxy1 proxy listen=tcp:10.83.17.189:60611 connect=tcp:10.246.13.238:3389 bind=host
用远程桌面进行连接。
桌面安装完成
将容器母本alpha制作成镜像，方便克隆多个容器给朋友使用。
停止容器
sudo lxc stop alpha
确认容器状态，已停止
ssme@ssme-server:~$ sudo lxc list +-------+---------+------+------+-----------+-----------+ | NAME | STATE | IPV4 | IPV6 | TYPE | SNAPSHOTS | +-------+---------+------+------+-----------+-----------+ | alpha | STOPPED | | | ConTAINER | 0 |
将test容器保存为ubuntudemo镜像
ssme@ssme-server:~$ sudo lxc publish alpha --alias ubuntudemo --public Instance published with fingerprint: 988a7ead01a8648ac1da593f306047ef926e327f1e7e1bca20fa5201b199299c
查看生成的镜像，大概有14个G
ssme@ssme-server:~$ sudo lxc image list [sudo] password for ssme: +------------+--------------+--------+--------------------------------------+--------------+-----------+------------+-----------------------------+ | ALIAS | FINGERPRINT | PUBLIC | DEscriptION | ARCHITECTURE | TYPE | SIZE | UPLOAD DATE | +------------+--------------+--------+--------------------------------------+--------------+-----------+------------+-----------------------------+ | ubuntudemo | 988a7ead01a8 | yes | Ubuntu bionic amd64 (20220228_07:42) | x86_64 | ConTAINER | 14020.81MB | Mar 2, 2022 at 7:07am (UTC) | +------------+--------------+--------+--------------------------------------+--------------+-----------+------------+-----------------------------+ | | 04d0148a8727 | no | Ubuntu bionic amd64 (20220228_07:42) | x86_64 | ConTAINER | 105.35MB | Mar 1, 2022 at 7:18am (UTC) |
再用镜像克隆出相应的容器，取名jellyfish
ssme@ssme-server:~$ sudo lxc launch 988a7ead01a8 jellyfish Creating jellyfish Starting jellyfish
克隆出来的容器jellyfish能够拥有alpha相同的内部环境，但是配置文件是空的，因此我们要给他配置
ssh：参考本章第四节远程桌面：参考本章第八节显卡：参考本章第六节
ssh
ssme@ssme-server:~$ sudo lxc config device add jellyfish proxy0 proxy listen=tcp:10.83.17.189:50601 connect=tcp:10.246.13.245:22 bind=host
远程桌面
ssme@ssme-server:~$ lxc config device add jellyfish proxy1 proxy listen=tcp:10.83.17.189:50611 connect=tcp:10.246.13.245:3389 bind=host
显卡
ssme@ssme-server:~$ lxc config device add jellyfish gpu0 gpu pci=0000:18:00.0
至此完成容器克隆
10.共享存储卷在容器alpha和jellyfish之间开辟一段共享存储卷。
创建共享共享存储卷，命名为dataset-vol
lxc storage volume create zfs-pool dataset-vol
设置存储卷大小
ssme@ssme-server:~$ lxc storage volume set zfs-pool dataset-vol size=300GB
连接自定义卷到alpha容器中，data是设备名，可以自定义，/data1是共享目录地址，连接前要先停止容器
lxc stop alpha lxc storage volume attach zfs-pool dataset-vol alpha data /data1
连接成功后不会有任何提示，直接进入容器，找到共享目录，创建一个文件夹做测试
sudo ssh ubuntu@10.83.17.189 -p 60601 cd /data1 sudo mkdir test
同理将共享存储卷连接到jellyfish
lxc stop jellyfish lxc storage volume attach zfs-pool dataset-vol jellyfish data /data1
然后进入jellyfish的/data1目录下，查看是否存在test文件夹
ubuntu@jellyfish:~$ sudo ls /data1 test
存在test文件夹，说明共享存储卷完成。
总结整一套服务器配置流程走下来因为走了很多弯路，增长了很多计算机知识，虽然起因是搞坏了服务器，但是感谢我的mentor没有放弃我，在整个过程中一直帮助我解决问题。
参考：https://github.com/shenuiuin/LXD_GPU_SERVER
