Presently, I am exploring the ELK (Elasticsearch, Logstash, Kibana) stack of tools. One of the Logstash scripts I am working on, is to parse a Java stack trace. In spite of multiple attempts and downloads from the Internet, I was not able to get the stupid thing working.
And then I read that the stupidity / gap lies in the implementation of Logstash, where the ‘file’ ‘input’ plugin apparently has issues handling the file path.
The plugin works without a flaw on Unix/Linux, but seems to have inconsistent behaviour on Windows. If your luck is good, the script will parse the input. If you try to push your luck and run the script once again, the script will simply refuse to read the data.
And this behaviour is not repeatable.