This repo has been updated to work on the older Gala Node v1.1.0 software. If you are looking to upgrade your existing nodes, check out the Upgrading the Container section below.
UPDATE for v2.x: Working on an update for the v2 node software. You can see the progress in the v2 branch.
- Overview
- Installation
- Building the Container
- Configuration
- Running the Container
- Checking Node Progress
- Upgrading the Container
- Frequently Asked Questions
This Dockerfile is used to run a Gala Node containerized in Docker. It uses a lean Debian 10 (Buster) image but should run on any Docker host that can run Linux containers (including Windows and MacOS).
If this guide is helpful to you, I'd appreciate if you used my Gala referal link if you decide to buy a node. Thanks, it really helps! 😃
- 1 CPU Core
- 256MB RAM (512MB Recommended)
- 10GB Disk Space (Solid State Recommended)
- IPv4 Address (IPv6 not supported)
- Stable Broadband Connection (50+ Mbit)
NOTE: These requirements are subject to change as the Gala Network grows. IPFS will be used in the near future, meaning these requirements will increase significantly from what they are now. Most important will be your bandwidth (namely upstream). Prepare accordingly.
Unfortunately, at the time of writing their Linux guide is fairly poor and not very scalable. It creates a systemd service and offers very little in terms of logging output and seeing the progress of your node throughout the day. For anyone who has tried to deploy a Linux node, it has been frustrating.
Docker solves both of these problems. Each container is run in an entirely isolated environment, allowing muliple nodes to easily be run on the same machine via the NODE_SPECIFIER variable. It also allows us to peek into the container and see the UI progress of each node. Since we no longer rely on systemd, we can run on a lot more Linux distros as well.
For more advanced setups, this could be run in a Docker Swarm cluster and be given CPU/memory limitations per container. As requirements grow, this will likely become more ideal for hosting multiple nodes across multiple VPS/machines but for now it is not required.
NOTE: The Gala Node does NOT support IPv6 networking! You must use IPv4 networking for it work properly.
You will first need to install Docker on your host machine (including VPS instances).
I personally find Vultr to be one of the most affordable VPS providers, feel free to signup with the link for a $50 $100 credit.
- Windows: Install Docker on Windows
- macOS:
- Intel Macs: Install Docker on Intel Macs
- M1 Macs: Install Docker on Apple Silicon
- Linux:
- Ubuntu: Install Docker on Ubuntu
- Debian: Install Docker on Debian
- Fedora: Install Docker on Fedora
- CentOS: Install Docker on CentOS
Gala node is currently only supported on the x86/64 architecture. Sorry, no Raspberry PIs or other ARMv7 devices yet!
Make sure you are running Linux containers, you can check here:
If it says "Switch to Windows containers..." that means you're already set to use Linux containers (which is good).
You will need to build the simple container on your Docker host machine. Once you build the container you can publish it to your own Docker container repository for reuse across installations.
$ git clone https://github.com/ewrogers/gala-docker.git
$ cd gala-docker
$ docker build -t gala-node:1.1.0 .
$ docker build -t gala-node:latest .
We run the build command twice so that we have both an explicit 1.1.0 version and a latest alias to that. This allows us to rollback to previous versions if something were to go wrong when using latest.
Verify the image was successfully built:
$ docker images | grep gala
> gala-node 1.1.0 7240420ada66 2 minutes ago 236MB
> gala-node latest 7240420ada66 2 minutes ago 236MB
Once you have Docker installed and built the container image, you will need to configure the nodes with your Gala credentials:
$ ./configure
If you have issues with the command above, try the following:
bash configure.sh or sh configure.sh depending on your shell.
You will be prompted to enter your Gala email and password for running the node. You can verify them by checking the .env file generated. You can either modify them there or re-run the configure script to overwrite them at any time.
If you still cannot run the script (or are on Windows), create the .env file manually with the following contents:
GALA_EMAIL=<your email login>
GALA_PASSWORD=<your password>
NODE_SPECIFIER=1
NOTE: You will need stop -> remove -> recreate any existing Gala node containers for them to take your new credentials (such as updating your password).
If you are trying to run nodes across multiple Gala accounts, you will need to create separate .env files for each one. For example, let's pretend I have a total of six nodes spread across three separate gala accounts (3x2).
Setting up each account credentials:
$ ./configure
$ mv .env .env.first
$ ./configure
$ mv .env .env.second
$ ./configure
$ mv .env .env.third
Since each Gala account expects a unique machine ID for their node(s), we'll generate one for each account instead of using the standard /etc/machine-id:
$ dbus-uuidgen > machine-id-first
$ dbus-uuidgen > machine-id-second
$ dbus-uuidgen > machine-id-third
Now you are ready to run the container! Here are the commands for running a single node on a single Gala account.
- If you have multiple nodes on the same account, see the Multiple Nodes section.
- If you have multiple Gala accounts with nodes, see the Nodes Across Gala Accounts section for how to run them.
$ docker run -itd --name "gala-node-1" \
--restart=unless-stopped \
--env-file .env \
-v /etc/machine-id:/etc/machine-id \
gala-node:latest
NOTE: The -v /etc/machine-id:/etc/machine-id line is very important! If you omit this, every time your container starts up it will have a different machine ID and you will lose progress across restarts! This mount ensures that you retain the same machine ID across reboots/restarts and always get proper credit.
You won't have an /etc/machine-id file locally. Instead you should run wmic csproduct get UUID and copy that UUID (without dashes) into a file called machine-id.txt and modify the command above to use -v machine-id.txt:/etc/machine-id instead.
$ docker run -itd --name "gala-node-1" --restart=unless-stopped --env-file .env -v machine-id.txt:/etc/machine-id gala-node:latest
If you want to run multiple nodes for the same Gala account, you will use the NODE_SPECIFIER variable for the second (and beyond) node for that account. To do that, update the NODE_SPECIFIER, append the -e NODE_SPECIFIER=x argument in the command above when running the additional nodes. This is not required for the first node, as it defaults to 1 automatically.
For example, on the second node for my account:
$ docker run -itd --name "gala-node-2" \
--restart=unless-stopped \
-e NODE_SPECIFIER=2 \
--env-file .env \
-v /etc/machine-id:/etc/machine-id \
gala-node:latest
You must also update the name of the container to be unique.
If you are trying to run nodes across multiple Gala accounts, you will need to create separate .env files for each one. For example, let's pretend I have a total of six nodes spread across three separate gala accounts (2x3).
This assumes you've follow the steps from the Configuration section on multiple env files and machine IDs.
Now it's just a matter of starting up each container:
$ docker run -itd --name "gala-node-first-1" \
--restart=unless-stopped \
-e NODE_SPECIFIER=1 \
--env-file .env.first \
-v machine-id-first:/etc/machine-id \
gala-node:latest
$ docker run -itd --name "gala-node-first-2" \
--restart=unless-stopped \
-e NODE_SPECIFIER=2 \
--env-file .env.first \
-v machine-id-first:/etc/machine-id \
gala-node:latest
Repeat the following while changing the --name, -e NODE_SPECIFIER=, --env-file= and -v machine-id-xxx:/etc/machine-id values as necessary. You can do this for as many accounts and nodes as you need per account. If you are just running a single node for an account, you can leave off the NODE_SPECIFIER argument, as the default of 1 will suffice.
You may be wondering why we specified the -it command line arguments while also using the -d (daemon) background flag. This gives an interactive TTY terminal that we can use to "peek" and see how the node is running.
$ docker attach gala-node-1
This should show you a simple terminal-based UI of the node's progress. Press Ctrl + P, Q to dettach from the container's TTY.
NOTE: Despite the program displaying "ESC to exit", you should not use that as it will terminate the node. Fortunately, it will automatically be restarted (assuming you used the --restart=unless-stopped argument when creating the container). Not a big deal, just avoid restarting your node unncessarily, but you won't lose progress for the day.
If you already have existing containers running the older Gala node software, it is easy to upgrade your nodes.
First run the steps from the Building the Container section. If you already have the repo cloned you can use git pull to just get the latest files, no need to clone again.
These commands will stop and remove the running container so it can be replaced.
$ docker stop gala-node-1
$ docker rm gala-node-1
Then it is just a matter of recreating the containers using the same commands as you did the first time around. See Running the Container for details.
Repeat for every node you need to upgrade. You should not have to re-create the .env or machine ID files.
You cannot vote using the Linux headless nodes; however, you can download the desktop node for Windows/macOS and vote that way. The UI allows you to allocate multiple votes based on the number of node licenses in your account. No need to touch your headless nodes, just open the UI vote and close the node software on your PC. The headless nodes running will continue to work without any interruptions.
Your Gala node will contact the Gala server periodically throughout the day. Each time it does, you gain credit for the day by unique machine ID. This is why is important that you retain the same machine ID each time your node runs, so you don't split up credit across several IDs. To get credit for the daily distribution, you must reach 100% for that machine ID. Meaning, if you get 95% on one and 5% on another, you do not receive credit.
This is only used when you are running more than one Gala node on the same machine (or VPS). By default, this value is 1, and is appended to your machine ID when contacting the Gala server for daily distribution credit. Since each machine has its own unique ID, you do not need it when running a single node on that machine.
However, when you are running multiple nodes on a single machine you should increment this for each node. For example, the second one would be -e NODE_SPECIFIER=2, and so on. This way you can get multiple node credit for the same machine ID without needing to spin up multiple VPS instances or physical machines.
First step should be to check your logs, via docker logs <container>. You may see errors regarding account authentication or network problems here. Next step would be to check your .env file that you have the correct and up to date credentials. Remember to stop -> remove -> recreate any containers if you recently updated this file.
You should check your Account Page under Node Info and see the expected number of nodes online. If the number remains incorrect after 5 minutes, check your containers for any errors (see next question).
The problem is likely that your machine ID is not unique in the Gala network. This is common when using Linux on VPS providers, but can be fixed easily.
Try the following in a root shell:
# rm /etc/machine-id /var/lib/dbus/machine-id
# dbus-uuidgen > /etc/machine-id
# cp /etc/machine-id /var/lib/dbus/machine-id
You will also need to restart any running Gala node Docker containers for the machine ID changes to be seen by them.