Update Steem Price Feed for Witness using CoinGecko API and WebSocket.

• Using CoinGecko websocket to update Steem price feed on the fly.
• In case websocket disconnected use CoinGecko REST API /simple/price endpoint to update price and restart websocket.
• Using Docker or PM2 approach to respawn the script on error.
• Customizable env file for setting up websocket, price time out, and sensitivity.

• Node.js (v8+)
• Makefile installed (optional)
• Docker (optional)

General step

• git clone this repo git clone git@github.com:coingecko/steemprice-feed.git
• cd steemprice-feed
• edit .env.sample to .env (cp .env.sample .env then nano .env or vi .env)

ACTIVEKEY= # Steem Active Key (String)
WITNESS= # Steem Witness Username (String)
PEGMULTI=1 # Peg Multiple settings (Number) [default: 1]
WEBSOCKETTIMEOUT=10 # Time out in minutes for checking web socket (Number) [default: 10]
PRICETIMEOUT=5 # Time out in minutes for last price check (Number) [default: 5]
SENSITIVITY=0.000 # Sensitivity for the price update. (Float) [default: 0.000]

• The default SENSITIVITY is set to 0.000 so that whenever there's a price change, the script will update STEEM price.
• WEBSOCKETTIMEOUT is to check whether the websocket update the price or not (using js setInterval)
• PRICETIMEOUT will check the last price update date time, to check whether web socket did get the price from the server. If not, restart the web socket connection.

• Make sure docker is installed. Installation guide

docker-compose build # Build the docker image
docker-compose up -d # Run the docker
docker-compose logs # to see log message
docker-compose logs --tail="all" # To see all logs
docker-compose down # To shut down docker

• Make sure Makefile are installed. Read Makefile for the full command being used.
• run make init to install pm2 (skip this step if you already install pm2, and run npm install)
• run make start (start the tool with pm2)

There is other command available as well:

• To stop the price feed update, run make stop
• To delete existing process running run make delete
• To check log message, run make log
• To check log message (last 1000), run make log-1000

• Fallback to Exchange data source.
• Handle error properly.

MIT