BUG : Info log level has a lower priority than debug

[lox]

We are creating our client like this:

    analyticsClient = new RudderAnalytics(process.env.RUDDERSTACK_WRITE_KEY ?? '', {
      dataPlaneUrl: process.env.RUDDERSTACK_DATA_PLANE_URL,
      logLevel: 'info',
      flushAt: 1,
    })

Still seeing log messages like:

[Rudder] debug: flushAt reached, trying flush...
[Rudder] debug: in flush
[Rudder] debug: in flush

We are running: "@rudderstack/rudder-sdk-node": "^2.0.4",

[MoumitaM]

@lox , Apologies for the delay, I will look into it.

[MoumitaM]

@lox, Currently in our SDK log level info has a lower priority than debug. So all the higher priority errors along with info will be printed.
Do you think that's the concern?

[lox]

I'm not sure I understand, are you saying that info < debug on your logger levels? If so, that seems very odd. 🤔

[lox]

Oh wow, yeah you are right:

rudder-sdk-node/src/Logger.js

Lines 2 to 9 in 48a9c6c

	const LOG_LEVEL_MAP = {
	log: 0,
	info: 1,
	debug: 2,
	warn: 3,
	error: 4,
	none: 5,
	};

Looks like it was introduced by @bardisg in 3619247 when winston was extracted. I'd hazard a guess that is a mistake / typo, as the more conventional logger levels from winston are:

https://github.com/winstonjs/winston#logging

const levels = {
  error: 0,
  warn: 1,
  info: 2,
  http: 3,
  verbose: 4,
  debug: 5,
  silly: 6
};

[lox]

I updated the title to reflect the issue better!

[MoumitaM]

Thanks, @lox for the clarification. We will fix this.

[MoumitaM]

@lox, This issue has been fixed and released. Check out 2.0.6