I have a doubt about the model bounds, I have a few services that only have request in periodic moments and although I think that is logic, when this service receive requests it show me that is critical.
I attach an image with the information but is there any option to manage these cases?
it seems that most of your signals consist of zeros, which are irrelevant and confuse the anomaly detector. It thinks that always having zeros is the expected behavior.
Consider using *_non_zoro_* functions in your job configuration. See API documentation for reference.
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.