Provides alerts in Discord when bonds behave abnormally.
The scope will be expanded in the future to support different metrics.
NOTE: this is currently not in use.
- A simple function is deployed on Cloudflare Workers to fetch bond snapshots and check that they are within set bounds. If out of bounds, a Discord message is sent.
- UptimeRobot calls the function periodically (every minute, at the moment).
- Pure Javascript/Typescript
- Simple APIs and environment
Initially, the inbuilt Cloudflare Workers cron trigger was used. However, it required some sacrifices:
- No response and hence no status code, which means we can't monitor errors
- Scheduled workers behave differently and aren't documented well
The StatusCake uptime service is used instead to perform an HTTP request against the Cloudflare Workers URL (in the format https://.workers.dev/), which triggers the check. If the Cloudflare Workers URL takes too long to respond, or returns a status that is not 200, the alerts channel in the Discord work server will receive a message.
The wrangler.toml file defines two deployments in Cloudflare Workers:
- Default deployment, used for development and named
discord-alerts - Production deployment, named
discord-alerts-production
There is a separate key-value store binding for each, and secrets are also separate.
To deploy to staging, run: yarn deploy
To deploy to production, run: yarn deploy:production
- The following secrets are required in order to run:
- WEBHOOK_URL - a Discord webhook URL
- Set this using the following command:
echo <URL> | yarn wrangler secret put WEBHOOK_URL- To set the variable in production, append:
--env production
- To set the variable in production, append:
- Copy the
.dev.vars.samplefile to.dev.varsand fill in the variable(s). - Run
yarn startto run the worker function locally - Run
yarn triggerin a new terminal to trigger accessing the worker function, and watch the output in the terminal thatyarn startwas run in.
StatusCake:
Cloudflare Workers: