The problem could be the "data timerange mode" option in the "Panel options" panel. Use the "Entire time range" option so the graph would be shown as the metric calculations of all the data on the date range chosen in the time picker. With the "Last value" option the graph will show only the data of the last bucket, that sometimes this could be empty
I'm already using the "Entire time range" option.
I do have something like 75 million documents in 24 hours though, is there a data fetching limit or a timeout that could be causing this issue?
Edit - Also I don't know if this makes any actual difference but I'm using a TSVB table not a graph.
Uhmm. Elasticsearch or Kibana logs reporting something? There is no limit, but in large request that an index can't handle usally appears a notification with the timeout error in Kibana.
Also we get a pretty dramatic spike in the CPU from 30% to 90%.
Elasticsearch and Kibana are on the same server and they communicate on their default ports.
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.