Severity levels

When you're logging events in your application, not all messages are created equal. Some might be routine information, while others could be critical errors that need immediate attention. That's where severity levels come in. LogTape provides five severity levels to help you categorize your log messages effectively.

Five severity levels

LogTape uses the following severity levels, listed from lowest to highest severity:

1. Debug: Detailed information useful for diagnosing problems.
2. Information: General information about the application's operation.
3. Warning: An unexpected event that doesn't prevent the application from functioning.
4. Error: A significant problem that prevented a specific operation from being completed.
5. Fatal error: A critical error that causes the application to abort.

NOTE

LogTape currently does not support custom severity levels.

Let's break down when you might use each of these:

Debug

Use this level for detailed information that's mostly useful when diagnosing problems. Debug logs are typically not shown in production environments.

logger.debug("Database query took {elapsedMs}ms to execute.", { elapsedMs });

Information

This level is for general information about the application's operation.

logger.info("User {username} logged in successfully.", { username });

Warning

Use this when something unexpected happened, but the application can continue functioning. This level is often used for events that are close to causing errors.

logger.warn("API rate limit is close to exceeding, 95% of limit reached.");

Error

This level indicates a significant problem that prevented a specific operation from being completed. Use this for errors that need attention but don't necessarily cause the application to stop.

logger.error(
  "Failed to save user data to database.",
  { userId: "12345", error: err },
);

Fatal error

Use this for critical errors that cause the application to abort. Fatal errors are typically unrecoverable and require immediate attention.

logger.fatal("Unrecoverable error: Database connection lost.", { error });

Choosing the right level

When deciding which level to use, consider:

• The impact on the application: How severely does this event affect the application's operation?
• The urgency of response: How quickly does someone need to act on this information?
• The audience: Who needs to see this message? Developers? System administrators? End-users?

Configuring severity levels

This API is available since LogTape 0.8.0.

You can control which severity levels are logged in different parts of your application. For example:

await configure({
  // ... other configuration ...
  loggers: [
    {
      category: ["app"],
      lowestLevel: "info",  // This will log info and above
      sinks: ["console"],
    },
    {
      category: ["app", "database"],
      lowestLevel: "debug",  // This will log everything for database operations
      sinks: ["file"],
    }
  ]
});

This configuration will log all levels from "info" up for most of the app, but will include "debug" logs for database operations.

NOTE

The lowestLevel is applied to the logger itself, not to its sinks. In other words, the lowestLevel property determines which log records are emitted by the logger. For example, if the parent logger has a lowestLevel of "debug" with a sink "console", and the child logger has a lowestLevel of "info", the child logger still won't emit "debug" records to the "console" sink.

The lowestLevel property does not inherit from parent loggers, but it is "debug" by default for all loggers. If you want to make child loggers inherit the severity level from their parent logger, you can use the filters option instead.

If you want make child loggers inherit the severity level from their parent logger, you can use the filters option instead:

await configure({
  // ... other configuration ...
  filters: {
    infoAndAbove: "info",
  },
  loggers: [
    {
      category: ["app"],
      filters: ["infoAndAbove"],  // This will log info and above
    },
    {
      category: ["app", "database"],
      // This also logs info and above, because it inherits from the parent logger
    }
  ]
});

In this example, the database logger will inherit the aboveAndInfo filter from the parent logger, so it will log all levels from "info" up.

TIP

The filters option takes a map of filter names to FilterLike, where FilterLike is either a Filter function or a severity level string. The severity level string will be resolved to a Filter that filters log records with the specified severity level and above.

See also the Level filter section.

Comparing two severity levels

This API is available since LogTape 0.8.0.

You can compare two severity levels to see which one is more severe by using the compareLogLevel() function. Since this function returns a number where negative means the first argument is less severe, zero means they are equal, and positive means the first argument is more severe, you can use it with Array.sort() or Array.toSorted() to sort severity levels:

import { type LogLevel, compareLogLevel } from "@logtape/logtape";

const levels: LogLevel[] = ["info", "debug", "error", "warning", "fatal"];
levels.sort(compareLogLevel);
for (const level of levels) console.log(level);

The above code will output:

debug
info
warning
error
fatal

Best practices

1. Be consistent: Use levels consistently across your application.
2. Don't over-use lower levels: Too many debug or info logs can make it harder to find important information.
3. Include context: Especially for higher severity levels, include relevant data to help diagnose the issue.
4. Consider performance: Remember that logging, especially at lower levels, can impact performance in high-volume scenarios.

By using severity levels effectively, you can create logs that are informative, actionable, and easy to navigate. This will make debugging and monitoring your application much more manageable.