It's not about long queries, we have timeout for that, it's about "zombie" search, when the end-user has left the web site, but elasticsearch is still searching.
I must setup a monitoring to count how many ES queries are "abandoned" by Php, maybe it's very little, but I am curious to know if it's possible to optimize our ES cluster by stoping "zombie" searches.
I think the Task API's X-Opaque-Id header support is the most suitable mechanism for this purpose; if a value is supplied in the search call, it can be used to cancel the tasks with that value after client has aborted.
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.