Aki

Overall Plan#

1. Image file download address: proxmox VE
2. USB boot tool: Rufus or Ventoy, the latter is more recommended
3. Proxmox does not have a default operation page; you can control it via the web interface using the connection provided after system installation, defaulting to https://ip:8006/
4. Change source

Enterprise source

Copy

## Remove the built-in source
rm /etc/apt/sources.list.d/pve-enterprise.list

## Add the official non-subscription source (choose one between this and the domestic non-subscription source below)
echo 'deb http://download.proxmox.com/debian/pve bullseye pve-no-subscription' >> /etc/apt/sources.list.d/pve-no-subscription.list

## This is the domestic non-subscription source
echo 'deb http://mirrors.ustc.edu.cn/proxmox/debian/pve bullseye pve-no-subscription' >> /etc/apt/sources.list.d/pve-no-subscription.list

Software source

Copy

## Edit source file: /etc/apt/sources.list
cp /etc/apt/sources.list /etc/apt/sources.list.bak && vim /etc/apt/sources.list

## Delete all content in the file or comment it out with #, and add the following content
deb https://mirrors.aliyun.com/debian bullseye main contrib
deb https://mirrors.aliyun.com/debian bullseye-updates main contrib
deb https://mirrors.aliyun.com/debian-security bullseye-security main contrib

5. Multiple disks can be configured as raid5 to provide sufficient storage space for virtual machines while ensuring relatively reliable data storage.

At least 3 virtual machines need to be created:
157.20.20.1 --> jumpserver --> Bastion host for remote operation and maintenance
157.20.20.2 --> nfs --> NFS file sharing system
157.20.20.3 --> docker --> Dedicated container for deploying streaming media
If choosing the k8s method, at least 2 more virtual machines need to be virtualized:
157.20.20.3 --> k8s-master --> k8s master node
157.20.20.4 --> k8s-node01 --> k8s worker node
157.20.20.4 --> k8s-node02 --> k8s worker node

Among them, the jumpserver needs at least 4C8G of resources, while other virtual machines should allocate at least 2 cores, with memory allocated as needed, without special requirements. If it is a k8s environment, the master node needs at least 2C6G, and the node nodes need at least 2C4G.

The reason for creating multiple virtual machines: to divide space by function, avoiding all components becoming unavailable due to improper operations leading to server file damage. Once this happens, repair and maintenance will become exceptionally troublesome.

Creating Virtual Machines#

ISO images need to be uploaded to pve's local storage for installing the basic operating system.

Various basic image download addresses: CentOS7, Rocky Linux, Debian
Click "Create Virtual Machine" in the upper right corner and select the appropriate image for installation.

Installing Jumpserver on the Virtual Machine#

Simple Introduction
Jumpserver is an open-source bastion host product under Feizhi Cloud, which is much more convenient than various ssh and scp tools. Official website

Deployment

curl -sSL https://resource.fit2cloud.com/jumpserver/jumpserver/releases/latest/download/quick_start.sh | bash

████████████████████████████████████████ 100%
[Success]: download install script to /opt/jumpserver-installer-v3.2.0
[Info]: Start executing the installation script.
[Info]: In an automated script deployment, note the message prompts on the screen.
████████████████████████████████████████ 100%
[Success]: The Installation is Complete.

For more commands, you can enter jmsctl --help to view help information.

After installation, access it in the browser at http://157.20.20.1 and log in with the default account admin/admin.

Creating the First Asset
Add assets in Asset Management --> Asset List.

Create authorization rules for user asset authorization in Permission Management --> Asset Authorization.

After authorization is complete, you can enter the console in the upper right corner to log in to the virtual machine and perform other operations. The flexibility and convenience are much higher than ssh tools. At this point, the jumpserver deployment is complete.

Creating NFS Shared File Server#

Create a second Linux virtual machine and allocate sufficient disk space. Plan data space and media space in advance to facilitate management and avoid confusion.

Create media and storage folders in the system, respectively serving as storage space and media space, and mount the disks properly.

## Confirm the space to be mounted
$ lsblk
## Create media and storage folders
$ cd /mnt && mkdir media && mkdir storage
## Format and mount the disks to the folders for the operating system to access
## File system formats can be xfs or ext4, etc.
$ mkfs.xfs /dev/sdb
$ mkfs.xfs /dev/sdc
$ mount /dev/sdb /mnt/storage && mount /dev/sdc /mnt/storage
## Modify /etc/fstab to ensure automatic mounting at boot
$ vim /etc/fstab
/dev/sdb    /mnt/storage    xfs    defaults    0    2
/dev/sdc    /mnt/media   xfs    defaults    0    2

Deploy the NFS server, taking Rocky Linux as an example.

1. Install NFS server components:

$ yum install nfs-utils 

2. Set permissions for the directories to be shared:

$ chmod 777 /mnt/storage          
$ chmod 777 /mnt/media      

3. Configure NFS access permissions in the file /etc/exports. For example:

$ vim /etc/exports
/mnt/storage         157.20.20.0/24(rw,sync,no_subtree_check)
/mnt/media         157.20.20.0/24(rw,sync,no_subtree_check)
## This will provide read-write access to clients in the 157.20.20.0/24 subnet

4. Start the NFS service:

$ systemctl start nfs-server
$ systemctl enable nfs-server # Set to start automatically at boot

5. Open the NFS connection ports in the firewall:

$ firewall-cmd --permanent --add-service=nfs     
$ firewall-cmd --permanent --add-port=2049/tcp 
$ firewall-cmd --reload

6. Test the NFS configuration:

$ showmount -e localhost 
/mnt/storage         157.20.20.0/24
/mnt/media         157.20.20.0/24

Tip
If you modify the contents in exports, you need to re-export the exports as follows:

Copy

$ exportfs -r
$ systemctl restart nfs-server

### Docker Server
Create a third Linux virtual machine and deploy the NFS server, taking Rocky Linux as an example.
```bash
$ yum install -y yum-utils device-mapper-persistent-data lvm2
$ yum-config-manager --add-repo https://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo
$ yum makecache fast && yum -y install docker-ce
$ systemctl start docker && systemctl enable docker

Upload the docker-compose file downloaded from GitHub through the bastion host's file management. Download link
The uploaded file is located in /tmp by default.

$ cp /tmp/docker-compose-linux-x86_64 /usr/local/sbin/docker-compose
$ chmod +x /usr/local/sbin/docker-compose

After completion, you can check the versions of docker and docker-compose.

$ docker version
$ docker-compose version

Quick Deployment of k8s Environment#

Open-source project: KuboardSpray-github
This project can be directly deployed on the jumpserver virtual machine. Note that port 80 is already occupied, so you need to change the port when mapping.

docker run -d \
  --restart=unless-stopped \
  --name=kuboard-spray \
  -p 8081:80/tcp \
  -e TZ=Asia/Shanghai \
  -v /var/run/docker.sock:/var/run/docker.sock \
  -v ~/kuboard-spray-data:/data \
  eipwork/kuboard-spray:latest-amd64
  # If you can't grab this image, you can try this backup address:
  # swr.cn-east-2.myhuaweicloud.com/kuboard/kuboard-spray:latest-amd64

Access it via http://157.20.20.1:8081, with the default account admin/Kuboard123.
Using kuboard-spray, you can manage nodes from 157.20.20.3 to 157.20.20.6 (specifically see which virtual machines to use as k8s environment nodes).
It is worth noting that a single server can only simulate a k8s cluster for testing and learning purposes. A real k8s cluster should have multiple master nodes and worker nodes distributed across different physical servers to ensure that a single machine's downtime does not affect the cluster's operation.

For detailed installation and deployment references, see the kuboard-spray official documentation.
Optimization reference:
To be updated

Deployment#

Docker Method#

Mount the shared folder to the local machine.

## Confirm if it can be mounted
$ showmount -e 157.20.20.2
## Mount
$ cd /mnt && mkdir media && mount -t nfs 157.20.20.2:/mnt/media /mnt/media
$ mkdir storage && mount -t nfs 157.20.20.2:/mnt/storage /mnt/storage
## Set fstab for automatic mounting at boot as mentioned earlier

Deploy qbittorrent via docker.

docker run -d \
  --name=qbittorrent \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -e WEBUI_PORT=8080 \
  -p 8080:8080 \
  -p 49152:49152 \       
  -p 49152:49152/udp \   
  -v /mnt/storage/qb-config:/config \     
  -v /mnt/media/qb-downloads:/downloads \  
  --restart unless-stopped \
  lscr.io/linuxserver/qbittorrent:latest

The following points need to be noted in the above code:

1. The default port is 6881, which generally does not need to be changed. Most PT sites ban port 6881, so it can be changed to 49152 or others.
2. The path set in -v /mnt/storage/qb-config:/config must be created before using the command to deploy.
3. Replace the path /mnt/storage/qb-config in -v /mnt/storage/qb-config:/config with the desired qbittorrent configuration path. All modifications will be saved, and if the container crashes, as long as the configuration is not lost, restarting the container will suffice.
4. The same modification is needed for the /mnt/media/qb-downloads part in -v /mnt/media/qb-downloads:/downloads.

Tip
The key points during docker deployment are the ports and mount paths. Just find the image on docker hub and modify accordingly.
The subsequent text will not elaborate on the modification methods during docker deployment; the basic points are the same as those in this section.

k8s Deployment of qbittorrent#

Deploy qbittorrent through the kuboard panel.
Find a suitable namespace (you can manually create one) to create the workload.

Configure container information.

Mount storage.

Publish the application.

After saving, wait for the container logs to indicate success, and you can access it through the defined published port. For example, in the image above, port 30001 is mapped to the container's 8080 port, so you can access it via http://157.20.20.3:30001. The default account for qb is admin/adminadmin.

Basic Settings for qb#

In the upper navigation bar of the qb page, there is a small gear icon, which is the entry for qb settings.

1. Regarding the Download option, in the save management section, you need to set the part in the red box in the image below. The first setting is related to sonarr and radarr; when performing hard links through sonarr and radarr, the folder will be modified. Setting it to manual allows you to take over the download folder settings. The second function is to extract and back up the downloaded file's torrent. Usually, downloaded media files do not come with torrent files, so this setting can back up the torrent to a specified folder.

2. In the BitTorrent tab, if you are using the BT method to search for resources on sites like EZTV, RARGB, ACG.RIP, etc., checking "Enable DHT (decentralized network) to find more users" will yield better results. If you are using PT sites, you can uncheck it.

The role of the tracker server is to help users increase download speeds and search for more user nodes. Tracker servers can be copied and pasted from trackerslist.com.

3. In the Advanced tab, checking the two tracker-related options helps connect to more tracker servers, thus increasing speed.

Deployment#

This section only introduces the docker deployment method. The deployment methods and notes for jackett, sonarr, and radarr are the same as the qb deployment method mentioned earlier. The k8s deployment method refers to the qb deployment method.

Tip
Remember to create the corresponding folders.

jackett#

docker run -d \
  --name=jackett \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -e AUTO_UPDATE=true  \
  -e RUN_OPTS= \
  -p 9117:9117 \
  -v /mnt/storage/jackett-congfig:/config \
  -v /mnt/media:/downloads \
  --restart unless-stopped \
  lscr.io/linuxserver/jackett:latest

sonarr#

docker run -d \
  --name=sonarr \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -p 8989:8989 \
  -v /mnt/storage/sonarr-config:/config \
  -v /mnt/media/tv:/tv  \
  -v /mnt/media/anime:/anime  \
  -v /mnt/media/qb-downloads:/downloads  \
  --restart unless-stopped \
  lscr.io/linuxserver/sonarr:latest

radarr#

docker run -d \
  --name=radarr \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -p 7878:7878 \
  -v //mnt/media/radarr-config:/config \
  -v /mnt/media/movies:/movies  \
  -v /mnt/media/qb-downloads:/downloads  \
  --restart unless-stopped \
  lscr.io/linuxserver/radarr:latest

overseerr#

docker run -d \
  --name=overseerr \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -p 5055:5055 \
  -v /mnt/storage/overseerr-config:/config \
  --restart unless-stopped \
  lscr.io/linuxserver/overseerr:latest

Settings#

jackett#

Access jackett via http://ip:9117 and click Add indexer to add.

sonarr#

Access sonarr via http://ip:8989,

In settings --> media management, set the media folder monitoring. Sonarr will automatically hard link media files from the downloaded series folder to this location.

In settings --> profiles, set the default language.

In index, set the default indexer, where you will need the API key from jackett. The added subscription index format is Torznab.

In settings --> connect, set the media center, where you will need to obtain the API key from emby. This part can wait until emby is deployed.

In settings --> download clients, set the connection to the qb downloader and create the corresponding monitoring directory.

In settings --> metadata, set the metadata retrieval information.

Once the above settings are complete, you can set up searches and monitoring for series in the Series section. It will also notify qb to download when monitoring is added, and after downloading, it will automatically hard link to the target folder for scraping.

radarr#

Radarr's basic settings are similar to sonarr, but radarr can be set to Chinese.

Log in via http://ip:7878, and in settings --> UI, you can set the page to Chinese.

In media management, adding "" can set the renaming to Chinese directories.

Overseerr#

Overseerr is a beautiful and powerful media management tool, which can be regarded as a combination of radarr and sonarr, managing them through API keys. After deployment, you will enter the guided setup page, where the API keys used can be found in the settings --> general section of sonarr and radarr.

Important points to note:

1. This software relies on the plex server, so please deploy the plex server first.
2. In general settings, choose to display the language as Chinese and the region as all.
   
3. TV and anime can be added separately in the server.
   
4. Media scraping heavily relies on TMDB access, so please set up a global proxy, for example, through a soft router for proxy access. If the proxy is set within the virtual machine where overseerr is located, please add -e https_proxy=http://172.17.0.1:7890 during the docker deployment process; otherwise, overseerr will not be able to set the proxy.

Refer to the deployment methods for emby and plex on DockerHub for reference. Here are examples:

emby#

docker run -d \
  --name=emby \
  -e UID=1000 \
  -e GID=1000 \
  -e TZ=Asia/Shanghai \
  -p 8096:8096 \
  -p 8920:8920 \
  -v /mnt/storage/emby-config:/config \
  -v /mnt/storage/emby-lib:/lib \
  -v /mnt/media:/media  \
  --restart unless-stopped \
  lscr.io/linuxserver/emby:latest

plex#

docker run -d \
  --name plex \
  --network=host \
  --restart=always \
  -p 32400:32400/tcp \
  -p 8324:8324/tcp \
  -p 32469:32469/tcp \
  -p 1900:1900/udp \
  -p 32410:32410/udp \
  -p 32412:32412/udp \
  -p 32413:32413/udp \
  -p 32414:32414/udp \
  -e TZ="Asia/Shanghai" \
  -e PLEX_CLAIM="plex_clam_code" \
  -e ADVERTISE_IP="http://ip:32400/" \
  -e PUID=1000 \
  -e PGID=1000 \
  -v /mnt/storatge/plex/database:/config \
  -v /mnt/storatge/plex/transcode:/transcode \
  -v /mnt/media:/data \
  plexinc/pms-docker

The plex_clam code can be obtained from the page: https://www.plex.tv/zh/claim/

Tip
Plex is not open-source media library software; its account management requires registration on the official website, and the plex_clam_code is obtained through the account.

Deployment#

docker run -d \
  --name=bazarr \
  -e PUID=1000 \
  -e PGID=1000 \
  -e TZ=Asia/Shanghai \
  -p 6767:6767 \
  -v /mnt/storage/bazarr-config:/config \
  -v /mnt/media/movies:/movies \
  -v /mnt/media/tv:/tv \
  -v /mnt/media/anime:/anime \
  --restart unless-stopped \
  lscr.io/linuxserver/bazarr:latest

Settings#

The general settings in bazarr are basically the same as those in radarr and sonarr.

The difference is that there are three places where you need to set the API key. The API keys for sonarr and radarr can be obtained from their respective pages, while the subtitle download API key needs to be obtained by registering an account on the corresponding subtitle website. I am using 射手网 (伪) for this, with the registration link provided in the text.