Actionable Logging for Smoother Operation and Faster Recovery

An active web application has a lot of things going on. Just getting meaningful data out of the web log can be a challenge, but what about the application log? Readable logs are an important tool for operations staff when debugging production issues. Appropriate logging can mean the difference between a 20-minute outage and a 2-hour outage while esoteric error codes are deciphered or developers are contacted to investigate.

Actionable logging aims to clean out all the extra chatter that comes out of an application, whether the loglevel was incorrectly applied, or the development team really wasn’t thinking about how the system would behave in a potentially volatile production environment. Every piece of information in the production log should be useful to operations, as stewards of the product.

It’s impossible to address every possible error condition that can arise in a production environment, but there are dozens of scenarios where simply changing where in the code an error is logged from can improve the quality of logs and reduce recovery times. Additionally, the separation of metrics and statistics from the application log allows operations to manage towards performance trends without getting in the way of problem solving.