Practical Tech on ottercoconut's Blog

Synology NAS Proxy and Emby Configuration

Mon, 16 Feb 2026 11:12:15 +0800

Recently, when using v2rayA on my Synology NAS, I encountered an issue where incorrect time synchronization caused it to stop working properly, which in turn prevented Emby from scraping metadata. Although I don’t know the exact reason, it’s an old problem. The version was very old, and I was using a Docker image from a third-party application store; currently, Docker can’t pull images via direct connection either. Here is my experience with solving this:

Reinstall v2rayA
Configure the Docker version of Emby.

Reference Websites:

Linux Backup Installation Method - v2rayA

XTLS/Xray-core: Xray, Penetrates Everything. Also the best v2ray-core. Where the magic happens. An open platform for various uses.

Implementing Transparent Proxy on Synology - v2rayA

sjtuross/syno-iptables: Some missing iptables modules for Synology

Setting up Proxy for Container Manager (Docker) on Synology DSM 7.2 - CSDN Blog

v2rayA

According to the official v2rayA documentation, there is no specially adapted version for Synology, so we use the generic binary file for installation. Since we cannot pull files from GitHub over a direct connection, we download it on our computer and transfer it to the NAS. In addition, the installation steps below are actually slightly different from the official documentation.

Copy v2rayA and xray

1
2
3
4


cp v2raya_linux_x64 /usr/local/bin/v2raya
cp xray /usr/local/bin/
chmod +x /usr/local/bin/v2raya
chmod +x /usr/local/bin/xray

Copy the GFWList proxy rule files included with xray: geoip.dat and geosite.dat
1 2 3

cd /usr/local/share mkdir xray cp .../g* xray

Modify the service configuration file

1

sudo vi /etc/systemd/system/v2raya.service

The content was generated by ChatGPT combined with the official documentation:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16


[Unit]
Description=v2rayA Service
After=network.target

[Service]
Type=simple
User=root
Environment="V2RAYA_CONFIG=/usr/local/etc/v2raya"
Environment="V2RAYA_LOG_FILE=/tmp/v2raya.log"
Environment="XRAY_LOCATION_ASSET=/usr/local/share/xray"
ExecStart=/usr/local/bin/v2raya --passcheckroot
Restart=on-failure
LimitNOFILE=1000000

[Install]
WantedBy=multi-user.target

Activate v2rayA. According to the official documentation, run it as a service
1 2

sudo systemctl start v2raya sudo systemctl status v2raya
Configure it to start up automatically on boot. Theoretically, the configuration is complete after this step. Along the way, I also solved the issue of transparency proxy not working on Synology.
1 2

sudo systemctl enable v2raya sudo systemctl is-enabled v2raya

Activate Transparent Proxy

arch	kernel	iptables version	system model	platform version
apollolake	4.4.180+	v1.8.3	DS918+	7.0.1-42218
apollolake	4.4.59+	v1.6.0	DS918+	6.2.3-25426
broadwell	3.10.105	v1.6.0	DS3617xs	6.2.3-25426
bromolow	3.10.105	v1.6.0	DS3615xs	6.2.3-25426
geminilake	4.4.180+	v1.8.3	DS920+	7.1-42661
geminilake	4.4.302+	v1.8.3	DS220+	7.2-64570

Due to the characteristics of the Synology system, first determine the architecture based on your model, then select the appropriate files for your machine from the downloaded modules from the GitHub repository (for example, my DS224 is geminilake).

Upload the corresponding ko modules to /lib/modules/, and upload the corresponding so modules to /usr/lib/iptables/, and that’s it.

Run sudo -i and then run the following insmod commands to attempt loading the ko kernel modules. Because the modules depend on each other, they must be loaded in a specific order.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21


insmod /lib/modules/nfnetlink.ko
insmod /lib/modules/ip_set.ko
insmod /lib/modules/ip_set_hash_ip.ko
insmod /lib/modules/xt_set.ko
insmod /lib/modules/ip_set_hash_net.ko
insmod /lib/modules/xt_mark.ko
insmod /lib/modules/xt_connmark.ko
insmod /lib/modules/xt_comment.ko
insmod /lib/modules/xt_TPROXY.ko
insmod /lib/modules/xt_socket.ko
insmod /lib/modules/iptable_mangle.ko
insmod /lib/modules/textsearch.ko
insmod /lib/modules/ts_bm.ko
insmod /lib/modules/xt_string.ko

insmod /lib/modules/nf_nat_ipv6.ko
insmod /lib/modules/nf_nat_masquerade_ipv6.ko
insmod /lib/modules/ip6t_MASQUERADE.ko
insmod /lib/modules/ip6table_nat.ko
insmod /lib/modules/ip6table_raw.ko
insmod /lib/modules/ip6table_mangle.ko

However, after a system reboot, the modules need to be reloaded, so generate a script with the above content at /usr/local/bin/load_v2raya_mods.sh.

Then restart the v2rayA service:

1

sudo systemctl restart v2raya

Emby

After configuring v2rayA, even though I used the method of modifying the configuration file, Emby’s traffic still couldn’t successfully pass through v2rayA. After various fruitless attempts to modify the IP configuration, it naturally occurred to me that this was caused by Docker running as a host service.

After modifying the configuration on v2rayA multiple times, I finally found that enabling the transparent proxy solved the problem, and it also fixed the issue of being unable to connect to the Docker repository. But a bigger problem arose: enabling the transparent proxy caused all external network access to fail (fortunately, local network connections still worked). I think “transparent proxy” can be understood as a “global proxy”.

For the issue of setting up a proxy specifically for Docker (Container Manager), please refer to the links in the preface.

Select Emby in “Images” and run it. Add the path to the media library in “Volume Settings”, add three environment variables in “Environment”, and select host mode in “Network”:
1 2 3

HTTP_PROXY = http://127.0.0.1:20171 HTTPS_PROXY = http://127.0.0.1:20171 NO_PROXY = localhost,127.0.0.1
Complete the basic setup, install the previous anime-related plugins, and the scraping was successful.

How to Use Host Proxy on WSL2

Fri, 30 Jan 2026 16:31:20 +0800

Since I’ve been frequently downloading various models with Python recently, I finally decided to solve the long-standing issue of WSL2 not being able to use the Windows host’s proxy. After wasting half an afternoon, I finally got it working. With the help of my capable assistant Gemini, the main steps are as follows:

.wslconfig settings in Windows
Core settings of v2rayN
Firewall settings in Windows
Cleaning up old settings in WSL2
~/.bashrc settings in WSL2
curl -v testing in WSL2

References:

Advanced settings configuration in WSL | Microsoft Learn

WSL2 使用 V2RayN 局域网 proxychains 代理方案 · Issue #2653 · 2dust/v2rayN

记一次用wsl2中共享宿主机的代理-v2rayN - 沉迷于学习，无法自拔^_^

.wslconfig

Access your personal account folder in Windows. Press Win + R, enter %UserProfile%, and hit Enter.
Check if there is a .wslconfig file. If not, create a new text file and name it .wslconfig.

Paste the following content into .wslconfig:

1
2
3
4
5
6
7


[wsl2]
# Enable mirrored networking mode
networkingMode=mirrored
# Allow WSL2 to access Windows localhost
localhostForwarding=true
# Automatically synchronize proxy settings (optional: true/false)
autoProxy=true

Note that the autoProxy parameter determines how WSL2 handles the proxy. Setting it to true means you don’t need to configure ~/.bashrc anymore, but the problems are:

When we use the env | grep -i proxy command, we will see many strange network-related variables, even though the proxy does successfully work.
The proxy cannot be toggled on or off within WSL2, causing a lot of traffic to go through the proxy unnecessarily.

Later, we will set it to false so that we can easily control the proxy switch inside WSL2, ensuring a clear and transparent WSL2 system.

In the Windows terminal, enter wsl --shutdown to shut down WSL2.

v2rayN

In the basic settings of v2rayN (version V7.15.7 at the time of writing), enable “Allow connections from the LAN” and “Open a new port for the LAN” (optional).
At the bottom left of the v2rayN client’s main interface, you can see the port open for the internet, which is 10810 in my case.
The system proxy here is “Set system proxy automatically”, and the routing mode is “Bypass (Whitelist)”.
Then select a node and keep v2rayN running.

FireWall

Type “firewall” in the Windows search box and select Windows Defender Firewall.
Click “Allow an app or feature through Windows Defender Firewall”.
Find or add v2rayN.exe and its core programs (such as v2ray.exe or xray.exe), ensuring that both Private and Public are checked. The usual paths are under v2rayN\ and v2rayN\bin\xray, respectively.
In the advanced settings of the firewall, create a new inbound rule to allow TCP traffic on port 10810 (corresponding to v2rayN).

wsl2

Cleaning Up Old Settings

1
2
3
4
5
6
7
8


# 1. Clear old, messy proxy environment variables
unset http_proxy
unset https_proxy
unset no_proxy

# 2. Re-set correct proxy (pointing to localhost port in mirrored mode)
export http_proxy="http://127.0.0.1:10810"
export https_proxy="http://127.0.0.1:10810"

Ensure that you have enabled mirrored networking mode in the first step and successfully restarted WSL2. Of course, this step is not strictly necessary either, because ~/.bashrc will automatically handle these old variables later.

~/.bashrc

1
2


# Same applies to other editors
sudo vim ~/.bashrc

After multiple experiments with Gemini, here is the content to append to the bottom of ~/.bashrc:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22


function proxy_on() {
 # Thoroughly clean up any residual variables (prevent conflicts from mixed case)
 unset http_proxy https_proxy ALL_PROXY NO_PROXY HTTP_PROXY HTTPS_PROXY all_proxy no_proxy
 
 # Set the port you verified works (since 10810 was tested successfully, use 10810)
 export hostip="127.0.0.1"
 export port="10810"
 
 export http_proxy="http://$hostip:$port"
 export https_proxy="http://$hostip:$port"
 export all_proxy="socks5://$hostip:$port"
 
 # no_proxy here only keeps localhost
 export no_proxy="localhost,127.0.0.1"
 
 echo "WSL Proxy: ON (127.0.0.1:10810)"
}

function proxy_off() {
 unset http_proxy https_proxy ALL_PROXY NO_PROXY HTTP_PROXY HTTPS_PROXY all_proxy no_proxy
 echo "WSL Proxy: OFF"
}

After saving the above content, enter:

1

source ~/.bashrc

And then:

1

curl -I https://www.google.com

The following output should appear, indicating a successful connection:

I also tested it with ping, which didn’t work but doesn’t affect its usability.

1

HTTP/1.1 200 Connection established

I tried downloading the model again, and it successfully worked:

1
2


import gensim.downloader as api
wv_from_bin = api.load("glove-wiki-gigaword-200")

Solution If It Still Fails

The most effective method is, after completing the above steps, to enter the following in WSL2:

1

curl -v https://www.google.com

Then feed the output result to an AI, and it will tell you what to do.

Deploying a Minecraft Server on Linux (CentOS)

Wed, 19 Mar 2025 18:48:59 +0800

I recently set up a physical server and wanted to use it to host a Minecraft server. After staying up late researching, I succeeded and am sharing this guide in hopes that it helps everyone, especially Linux users without a public IP environment (like me).

Reference Websites

Installing and using Red Hat build of OpenJDK 21 on RHEL | Red Hat Product Documentation

SakuraFrp Launcher Installation / Usage Guide | SakuraFrp Documentation

CentOS | Docker Docs

Linux Terminal Server Hosting Tutorial ★ No Panel ★ Minecraft_bilibili

From the Bilibili uploader 翱翔大使, which is the main source of ideas for this whole article.

Java Configuration

Running Minecraft requires the corresponding version of the Java environment; here I installed OpenJDK 21.

1
2


sudo yum install java-21-openjdk
java -version // verify if successfully installed

If the server has multiple Java versions, you can use alternatives to switch versions.

1

alternatives --config java

As shown below, we enter 2 and press Enter to switch to the required version.

Game Deployment

First, download the Minecraft server software from the following URL. Here I downloaded Banner (1.20.1), which supports Fabric.

[MohistMC](MohistMC - Home)

Once downloaded, you will get a file like banner-1.20.1-800-server.jar. Next, open your SSH client to operate on the server:

1
2
3


cd /home/username // switch to personal directory or desired installation location
mkdir Minecraft // create a folder for the game
cd Minecraft

Using the SFTP feature in your SSH client (or any other file transfer method), copy the game file banner-1.20.1-800-server.jar you just downloaded into the newly created /home/username/Minecraft folder.

Next, let’s write a startup script for the server.

1

nano start.sh

Fill in the following content, but note the purpose of each parameter:

-Xmx is the maximum allocated memory, -Xms is the minimum allocated memory. I have 32GB of memory and allocated 6GB to the game (feel free to allocate more).
banner-1.20.1-800-server.jar is the name of the game file you just downloaded.

1
2


java -Xmx6144M -Xms6144m -jar banner-1.20.1-800-server.jar
stty echo

Press Ctrl + O to write, Enter to confirm, and Ctrl + X to exit.

Next, grant execution permissions to start.sh to avoid permission denied issues.

1

chmod 777 start.sh

Then install screen. Simply put, screen is a tool that helps users create independent sessions that can be resumed at any time.

1

yum install screen

screen has the following common commands:

1
2
3


screen -S [name] // create a new screen named "name"
screen -ls // list names and ports of all running screens
screen -r [port] // attach to the screen with the specified port

Next, create a new screen to run the script.

1

screen -S Minecraft

In the newly appeared session, run start.sh.

1

./start.sh

Then everything should go smoothly. I didn’t encounter any errors here. Finally, it comes to ...EULA... asking us to agree to the EULA. Enter true and press Enter. After a short wait, the game server will run successfully on port 25565.

To leave this Minecraft screen, just press Ctrl+A+D.

For game rule changes (like ‘whether cracked players are allowed to join’), you need to modify the contents of server.properties.

Regarding connection: If it’s a cloud server, map the port in the administration panel, then connect to domain:port in your Minecraft client.

However, for a physical server like mine, or a Linux device, or a personal PC without a public IP, we need to proceed with Intranet Penetration.

Intranet Penetration

Here I am using SakuraFrp, which is well-known and reliable in the Minecraft community, for intranet penetration. Other tools are mostly similar.

Docker

SakuraFrp on Linux runs on Docker, so let’s deploy Docker first. The operations entirely follow the official documentation.

1
2


sudo dnf -y install dnf-plugins-core
sudo dnf config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo

I encountered very slow installation speeds and download failures here. Running the command again solved the problem.

1

sudo dnf install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

After the installation is complete:

1

sudo systemctl enable --now docker

1

sudo docker run hello-world

The run hello-world test above is very likely to fail. Let’s solve this issue by referring to the following two articles:

[Complete Solution] Failed to run hello-world image after Docker installation: Unable to find image ‘hello-world:latest’ locally - CSDN Blog

Failed or Timeout Running hello-world Image in Docker - Paul7777 - cnblogs

Combining the two above will eventually solve the problem. First, let’s configure the daemon file.

1

nano /etc/docker/daemon.json

Copy the following content into it:

At the time of my testing (2025/3/19), the following mirrors are still working.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14


{
 "registry-mirrors": [
 "https://h59pkpv6.mirror.aliyuncs.com",
 "https://registry.docker-cn.com",
 "https://docker.mirrors.ustc.edu.cn",
 "https://hub-mirror.c.163.com",
 "https://mirror.baidubce.com",
 "https://do.nark.eu.org",
 "https://dc.j8.work",
 "https://docker.m.daocloud.io",
 "https://dockerproxy.com",
 "https://docker.nju.edu.cn"
]
}

Save + Exit. Next, restart docker, and execute the test once more.

1
2
3


sudo systemctl daemon-reload
sudo systemctl restart docker
docker run hello-world

I successfully installed docker up to this point. If the test still fails, please check the content of daemon.json to see if there are missing or extra commas and brackets.

SakuraFrp

For deploying SakuraFrp on Linux, the official documentation provides a detailed solution.

First, run the following command as an administrator in the terminal:

1

sudo bash -c ". <(curl -sSL https://doc.natfrp.com/launcher.sh)"

After installation, it should automatically output logs and prompt you to fill in the access token. This (or the subsequent operations) can be found in the management panel on the SakuraFrp official website.

After logging in, you will be able to see its log files. Below are the common operations to start it and view logs:

1
2


docker start natfrp-service
docker logs natfrp-service

As shown below, we now need to physically operate the server.

Open a browser (Linux usually comes with Firefox) and access the URL after “Usage” to open the WebUI.

Then you will see there is nothing under “Tunnels”, only a plus sign. At this time, we enter the SakuraFrp management panel, find Tunnel List under Services, and create two new tunnels as shown below:

The first one with port 7102 is the WebUI for SakuraFrp on the server, intended for remote management.

The second one with port 25565 is for the Minecraft server.

Return to the WebUI interface and refresh, you will see the two tunnels just created. Double-click them respectively, and then go back to the terminal log interface.

The links in red characters as shown in the image are the remote access links for the WebUI and Minecraft. Just copy the Minecraft one into the game, and you can connect to it.

Conclusion

We are done! (A screenshot of my server’s spawn point >w<)