FAQ: Understanding LogScale Log Error Levels

Last Updated: 2021-12-17

LogScale generates a number of logs, with each event in the logs identified with a specific log-level. These log levels can help you to identify the severity of an issue:

  • INFO

    General information about the system and execution.

    This serves to provide us with snapshots of what is happening in LogScale, and where the server may be during execution. Anything logged as info should always be the correct behaviour and not an error.

  • WARN

    This level is a warning about behaviour that goes against intended use, but is not erroneous per se.

    For instance, invalid input for a query function such as invalid arguments of arguments of the wrong type be highlighted as indicating a failure to execute correctly, but does indicate that the system has failed or that LogScale itself has failed.

    These events are more about loss of precision of the output, faulty user input, and so on.

  • ERROR

    This level should be reserved for internal errors.

    This is the level on which LogScale fails and execution cannot proceed. LogScale may have got to a bad state is unable to continue or from which the server is nable to recover easily recover.

    For an error, LogScale has no choice but to stop execution and start over.