Add metrics for log events

[StephenButtolph]

Different log levels are frequently not enabled when running in production due to processing costs. However, significant changes in the frequency of logs is unexpected. To enable monitoring of disabled logs, it would be nice to add a metric that records how frequently different log levels would be logged.

This interface defines the different log levels:

avalanchego/utils/logging/logger.go

Lines 21 to 41 in 318da00

	// Log that a fatal error has occurred. The program should likely exit soon
	// after this is called
	Fatal(msg string, fields ...zap.Field)
	// Log that an error has occurred. The program should be able to recover
	// from this error
	Error(msg string, fields ...zap.Field)
	// Log that an event has occurred that may indicate a future error or
	// vulnerability
	Warn(msg string, fields ...zap.Field)
	// Log an event that may be useful for a user to see to measure the progress
	// of the protocol
	Info(msg string, fields ...zap.Field)
	// Log an event that may be useful for understanding the order of the
	// execution of the protocol
	Trace(msg string, fields ...zap.Field)
	// Log an event that may be useful for a programmer to see when debuging the
	// execution of the protocol
	Debug(msg string, fields ...zap.Field)
	// Log extremely detailed events that can be useful for inspecting every
	// aspect of the program
	Verbo(msg string, fields ...zap.Field)

It would be nice to add metrics here:

avalanchego/utils/logging/log.go

Lines 82 to 87 in 318da00

	// Should only be called from [Level] functions.
	func (l *log) log(level Level, msg string, fields ...zap.Field) {
	if ce := l.internalLogger.Check(zapcore.Level(level), msg); ce != nil {
	ce.Write(fields...)
	}
	}

that would count the number of calls by loggerName and log level

[github-actions]

This issue has become stale because it has been open 60 days with no activity. Adding the lifecycle/frozen label will cause this issue to ignore lifecycle events.