Logging

You don't always have access to the debugger. When you're in production, you still need to know what's going on with your app. That's where logging comes in.

Log Levels

Sometimes we just want the CliffsNotes, while other times we want everything. So, to handle this, we have multiple log levels:

Fatal: The Most Severe Log

• Causes the application to crash
• Failing in a way that is unrecoverable

Error: Something Bad Happened

• Typically causes whatever operation/action you're doing to fail

Warning: Something Unusual Happened

• A condition that we might want to be aware of happened, but is not indicative of a failure
• Warning would not indicate that something did not work, but it could be an interesting clue
• Less severe than Error

Info: Just Thought You Should Know

• An object was successfully saved
• User logged in
• User logged out
• Informational about the activity that was happening

Debug/Trace: Extremely Verbose

• Informational only -- not indicative of anything bad happening
• Informational to help you figure things out
• Should be used in any part of the application that is rapidly repeatable

Rolling Logs

Usually, you and your team don't need an endless history of logs. The logs from 5 years ago would be pretty irrelevant to your team. So, usually logs are rolled, which means that only a specific set of logs over a given time period are kept. Many people use a 5-7 day rolling log:

• Log.0.txt (Sunday)
• Log.1.txt
• Log.2.txt
• Log.3.txt
• Log.4.txt
• Log.5.txt
• Log.6.txt (Saturday)

Once you get back to Sunday, you start writing over your previous Sunday log.