Welcome to the Java Stack Digest filter plugin pages

This site introduces the utility of computing Java stack digests, highlights the stack digest stability challenge, and also provides a debugger to help you configure and tweak your Java Stack Digest filter plugin configuration.

Why generating stack digests?

A full Java stack trace is very helpful to analyse and debug a single error, but is not very handy to compare two errors.

The idea is to turn a full Java stack trace into a short and stable digest, that could help matching several distinct occurrences of the same type of error:

short for easing elasticsearch indexing, and take advantage of it (we use a MD5 hash),
stable is the tricky part, as the same type of error occurring twice may not generate exactly the same stack trace (see challenge).

This done, it becomes easy with elasticsearch or any other logs centralization and indexation system to:

count distinct type of errors that occur in your code over time,
count occurrences and frequency of a given type of error,
detect when a (new) type of error occurred for the first time (maybe linking this to a new version being deployed?).

The stack digest may also become a simple error id that you can link your bug tracker with...