I've been reviewing different ways to aggregate log messages together that have a start event but no end event. Been struggling with the logstash aggregate filter plugin and was looking at retrofitting an old entity-centric model for a previous version of elasticsearch Entity-Centric Indexing - Mark Harwood | Elastic Videos when I realized elasticsearch 7.13 transforms introduce the concept of 'latest' which negates my need for a bunch of external scripts (hopefully).
I am looking at the "Getting Web Session Details by using Scripted Metric Aggregation" sample painless script, which produces session details, including session duration. Because my logs do not have an end-time, I need to make use of a timeout interval, something like a 30 minute window for aggregating message events based on my group by.
Is this possible to do within the transform by adjusting that script and could anyone help?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.