Log output query

[kaihendry]

2021-09-21T01:30:02.820Z	undefined	WARN	{
    "level": "warn",
    "msg": "a warning!",
    "timestamp": "2021-09-21T01:30:02.820Z",
    "_tags": []
}

Why is there these extra fields like

• 2021-09-21T01:30:02.820Z - redundant since my logs already have a timestamp by AWS
• "undefined" - I'm really not sure what the purpose of this is
• WARN - this is already in my structure log, why is it here? make it easier to filter?

[KyleRoss]

• 2021-09-21T01:30:02.820Z - redundant since my logs already have a timestamp by AWS
  This timestamp field is not added by LambdaLog by default, it appears you have set metadata somewhere in your application that has added this field to the log message.
• "undefined" - I'm really not sure what the purpose of this is
  I'm not sure where the undefined is coming from. I've been using this project in several enterprise applications and haven't encountered this issue. Given that it's shown before the WARN in your example, I'd say it's something with CloudWatch as we cannot change anything before the WARN.
• WARN - this is already in my structure log, why is it here? make it easier to filter?
  Since AWS logs in CloudWatch are labeled by the console method used, not all log levels within LambdaLog (including custom ones) would be labeled properly in CloudWatch because there is not a corresponding console method. For example, if you have a custom log level of fatal, we couldn't use console.fatal() to log the message as it does not exist, in this case, we would use the next best thing, console.error(). This would label the log message with ERROR instead of fatal in AWS which wouldn't allow you to filter by fatal messages.