I have set up ILM policy on data stream and when I manually perform rollover trough rollover api, it's just a blink of an eye and new index is created.
BUT, if I use ILM, there is some step called check-rollover-ready and it takes very long time before actual rollover action. Is there some description of this check, or exists way to speed it up? It takes a very long time even if index has about 10 documents. I dont want to imagine what takes up at our bilions documents indices.
ILM only checks for and triggers rollover periodically and I think the default is every 10 minutes or so. This works great for large indices but can cause problems for small unrealistic tests. You can change the interval but I do not remember how of the top of my head. I would recommend running a more realistic test instead.
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.