Avoid these Things When Logging from Your Application

Editorial note: I originally wrote this post for the LogEntries blog.  You can check out the original here, at their site.  While you’re there, sign up for the service and check it out.

It seems almost strange to talk about avoiding things while logging.  After all, logging is your last line of defense or your salvation in many cases.  Some crazy bug in the field that shows up every third full moon?  An external auditor looking at your app’s runtime behavior?  Logging to the rescue.

So naturally, is stands to reason that you would want to log just about everything your application.  Whenever there’s any doubt, slam a logger call in there and let log level sort out the details.  You can always filter logs, but you can’t magic stuff into them after the fact.  So why, then, talk of avoidance?

Well, it turns out that, while logging may be a highly inclusive activity in terms of what should be included, there are ways to create problems.  You want to be liberal in terms of what you log, but judicious and wise in terms of how you log it.  You don’t want to indulge in a feckless free-for-all when it comes to the calls you make to your application’s logger.

So what are these problems, and how to avoid them?  Let’s take a look at some things that can come back to bite you.

Read More

Keeping Your Code Clean while Logging

Editorial Note: I originally wrote this post for the LogEntries blog.  You can check out the original here, at their site.  While you’re there, take a look at their product offering that handles your log aggregation, search, and alerting needs.

In my consultancy practice, one of the things that I do most frequently is help teams write so-called “clean code.”  Usually, this orients around test-driven development (TDD) and writing code that is easily maintained via regression tests and risk-free refactoring.  Teams want to understand how to do this, and how to do it in their production code (as opposed to in some kind of toy “let’s build a calculator” exercise).

One of the most prominent, early sticking points that rears its head tends to be application logging.  Why?  Unit testing is all about isolating objects, instantiating them, and rapidly verifying their behavior in memory.  Logging is all about dumping as much information as possible to a file (or database, service, etc with appenders) from as many places as possible.  Unstoppable force, meet immovable object.

For instance, consider the following C# code from a “let’s build a calculator” exercise.

[TestMethod]
public void Adding_Numbers_Works()
{
    var target = new Calculator();

    var result = target.Add(2, 2);

    Assert.AreEqual<int>(4, result);
}

public class Calculator
{
    public int Add(int x, int y)
    {
        return x + y;
    }
}

Life here is good.  The calculator’s add method returns the sum of the integers, and we have a test method that supplies 2 and 2, and confirms a result of 4.  That’s a screaming fast, very clear test.  But, consider what happens if we amend our calculator to take advantage of logging.

public class Calculator
{
    public int Add(int x, int y)
    {
        Logger.Log($"Entering add method with parameters {x} and {y}.");
        Logger.Log($"Returning result of {x + y}");
        return x + y;
    }
}

Now, with that same test, everything goes wrong.  Instead of passing, it throws an exception about not having write access to some weird directory on your machine that you’ve never even heard of.  After digging a bit, you see that it’s the directory in which the test runner executes when running your solution’s tests.  Hmmm.

Read More