Skip to content

The data and DB for Bracket Club

Notifications You must be signed in to change notification settings

bracketclub/data

Repository files navigation

data

The data and DB for bracket.club.

Build Status

Running each watcher

Watchers

  • entries
  • scores

Sports

  • ncaam
  • ncaaw
  • nba
  • nhl
  • wcm
  • wcw
node watchers/$WATCHER --sport=$SPORT --year=$YEAR

Via pm2

# Any command run through this npm run-script will use check the hostname==localhost
# to see if it should run in PRODUCTION mode. You can also pass NODE_ENV=production to force it
# They will also use the current year for $YEAR

# Start only one
npm start $WATCHER $SPORT

# Run other pm2 commnads
npm start $WATCHER $SPORT restart
npm start $WATCHER $SPORT stop
npm start $WATCHER $SPORT delete
npm start $WATCHER $SPORT logs

Before Starting Scores Watcher

For the score watcher to run properly, each of the teams from bracket-data should have a match against wherever the scores are fetched. This can be done inside bracket-data by cloning that repo and running the npm run names command. If there are any names missing, then the bracket-data package should be published and pulled into this project before npm start scores is run.

Integration Tests

# Connects to the DB and adds masters/entries/users to all sports every 5 seconds
npm run integration

So you missed a finished score in production

# Want you probably want to do if you missed something today.
# Since order of masters matters in the database the --teams param order
# is important. This will take the current bracket and look up the scores for today
# (or use --date) and apply each result that the teams played in.
NODE_ENV=production npm run insert-by-team -- --sport ncaaw --teams uconn "notre dame"

So you missed an entry tweet

NODE_ENV=production npm run find-entry -- --sport ncaam --year 2018 --tweet TWEET_ID

Dumping a database

# Connection string will come from NODE_ENV and it will save to sql/$ENV.sql
NODE_ENV=production npm run dump

Setting up environment variables

# Create a .env file with all the necessary env vars for production
echo "TWITTER_KEY=$TWITTER_KEY" >> .env
echo "TWITTER_SECRET=$TWITTER_SECRET" >> .env
echo "TWITTER_TOKEN=$TWITTER_TOKEN" >> .env
echo "TWITTER_TOKEN_SECRET=$TWITTER_TOKEN_SECRET" >> .env
echo "POSTGRES_URL=$POSTGRES_URL" >> .env

Production Database on Heroku

# Open a psql shell
npm run heroku:psql

# Import current local sql file to production
npm run heroku:psql < sql/production.sql

# Open webpage to heroku DB
npm run heroku:web

Linode CLI Commands

# Create or rebuild the linode based on the current stackscript
# Pass it a path to a public key in order to allow SSH access
# The .env file will be used to populate vars to the Linode
npm run deploy build

# Update the stackscript with whats in the repo
npm run deploy stackscript

# IP address of linode
npm run deploy ip

# Delete linode
npm run deploy delete

Rough Steps to Launch Each Year

For Entries

  • Update and publish bracket-data with new year and teams
  • npm install bracket-data@latest --save in web app and this repo
  • Update web app config with new year
  • Locally: npm run dump:prod-to-dev
  • Locally: node watchers/entries --sport=$SPORT --year=$YEAR
  • Start api and postgres locally
  • Run webapp and make sure everything renders
  • Deploy to Linode npm run deploy build
  • ssh into linode and run cd data; npm start entries ncaam
  • Locally test that logs work: ssh bc "cat ~/data/logs/pm2/entries-ncaam.log"
  • Run this once and then stop it as it seeds the first master bracket npm start scores ncaam

For Scores

  • Go to bracket data and check npm run names works to ensure that the names in bracket-data all match the scoreboard
  • Go to scores and check that the tests pass, especially the ones getting data directly from urls. If anything breaks, fix and write new tests
  • Try running npm start scores ncaam locally once the first games start
  • ssh into linode and run cd data; npm start scores ncaam