Yet-Another-Logger that pushes logs to log servers with axon to delegate network overhead

Yet Another Logger that pushes logs to log servers with axon/tcp to delegate network overhead.

If you're like us and you distribute logs to several remote services, you may be using a tool like Winston to do so, and while this is helpful it's also brittle to use in mission-critical applications.

YAL's solution is to simply distribute messages over TCP (via axon) to one or more log servers, delegating the task of processing or shipping messages off to these services. This is especially problematic when logging services are enabled only in production, and thus may not be tested properly locally or in staging environment. This can and likely will bite you.

Check out the YAL server.

$ npm install yal
  • through axon re-connection and failover is supported
  • when NODE_ENV is "development" the log level defaults to "debug"
  • when NODE_ENV is not "development" the log level defaults to "info"
  • when NODE_ENV is "test" the logger will not write to stdio
  • the log level may be altered via the LOG_LEVEL environment variable
  • standard syslog levels are supported: debug, info, warn, error, critical, alert, emergency
  • you may distribute messages to one or more axon log servers
  • YAL writes to stdio for local logging / debuggin
  • YAL sends a creation .timestamp
  • YAL sends the .hostname

The following example shows two servers, with the logger doing round-robin requests between the two.

var Logger = require('yal');
var log = new Logger([
  log.info('viewed page', { user: 'tobi' });
}, 300);
  log.info('signed in', { user: 'jane' });
}, 1000);
  log.error('oh no boom', { something: 'here' });
}, 3000);

Pass a string or array of addresses.

Send a log message.

Disable stdio.