TODO
This is a quick and dirty laundry list of ideas and improvements for future version of the Jahia Log Analysis tool :
- In Jahia, add sessionID output and support in the log analyzer for this output.
- Documentation with screenshots and maybe even a small movie to illustrate how to use
- Look into possibility to detect in which subsystem a thread is (Lucene, etc...), could we do this through stack analysis ? What about global Jahia actions ? (see ThreadLogic that even uses stack traces to issue recommendations based on certain common traces)
- Add support for parsing /site/ in URL
- Fix URL parsing issues to reflect modern Jahia DX URLs
- Make it possible to develop plugins for the line analyzers
- Make it possible to configure new line analyzers simply by using regexp -> property mapping in ES
- Make it possible to incrementally parse logs so that we can for example schedule the log analyzer in cron jobs to populate the ElasticSearch backend
- Fix a better solution for the configuration if we are to support plugins
- Make it possible to use scripts (Groovy,Javascript) to extend the log analyzer
- Transform it into an Apache Karaf OSGi application ?