adamcanray/Private-IPFS-Worker

IPFS Worker Node

This repository is for IPFS Worker Node.

Run

1. Build the image

   docker build . -t ipfs_worker --no-cache

   if we want to build the image for specific platform (ex: --platform linux/arm64/v8).

2. Run the container

   docker run --name <name> -e MANAGER_IPFS_ID=<manager_ipfs_id> -e MANAGER_IP_ADDRESS=<manager_ip_address> -d ipfs_worker

   we can get the MANAGER_IPFS_ID and MANAGER_IP_ADDRESS from the manager node:

   <!-- go to the manager node terminal -->
   docker exec -it <name> /bin/sh
   
   <!-- below command will show detail in json structure, just copy the `ID` -->
   ipfs id
   
   <!-- below command will show interface config, copy the ip from eth0 (or other interface) -->
   ifconfig

   if we want to expose the port, we can add for example -p 4001:4001 -p 5001:5001 -p 8080:8080 to the command, but on ipfs config we are not exposed the API server (:5001) to the public.

Addition

About project

This project is running well with alphine linux with arm base (linux/arm64/v8), if you run this project using different distribution, maybe you should adjust some scripts or service config (ex: services/ipfs, etc).

Porject Debugging

If the ipfs daemon is not running, we can start it manually for each node with rc-service ipfs start command. First, we should exec into the container

docker exec -it <name> /bin/sh

then start the ipfs daemon.

rc-status -a
rc-service ipfs start
touch /run/openrc/softlevel
rc-service ipfs restart
rc-update add ipfs default
sleep 1
rc-status -a

Notes

• In previous, manager and worker project is in one code base, we run with docker compose to simplify the automation, since currently the worker project is not on one code base anymore, we should run it in some steps (ex: running worker node manually, then run worker nodes and pointing it to the worker manually), see Run section above.
• The swarm.keyis should be confidential (generated once from manager).
• do step 2 on Run section above for running other worker nodes.