Logging and Crash Reporting Case Study

The Business Problem

When something went wrong, the server wrote the details to a single text file on its own hard drive. Nobody watched that file, it grew without limit and, in a bad week, could eventually fill the disk and take the system down. On the desktop, crashes were only known about if an adviser happened to report them.

Problems were found late, by users, and were hard to diagnose.

The Solution

wildesoft.net replaced the old log file with centralised, structured logging. The server now records errors and the time each job takes in one searchable place, and the desktop application automatically sends a report to the same place when a serious crash occurs, while showing the adviser a clear message.

The Result

CXCS now sees problems before its staff do. The new logs have already uncovered and fixed a hidden calculation error that had gone unnoticed, and slow jobs can be identified and improved. The risk of the server filling its own disk has gone completely.

Good visibility is the foundation of any modernisation – you cannot improve what you cannot see.

See all case studies for Nutricalc