I am experiencing several documents missing after the upgrade to 7.0. Metricbeat dashboard used to show a regular update of the cpu and memory , however now it now shows a dash (-) for 1m before showing update. this makes it very difficult to track the health of the servers.
is there any way to go back to how it was before to prevent showing no units. (picture below shows explanation )
I can't create proper watchers and keep track of when the cpu is high or in danger like this.
Just to be sure, I assume these dashboards were created using the metricbeat setup command, is that correct?
Could you also double check the logs for your beats to see if they're facing any errors and double check the clock on your servers?
One more thing to try would be to run a query using the Kibana console under the Dev Tools to see the raw documents themselves, so that you can see the field's names and values and compare that to the queries used by your dashboard.
Some incompatibility between a 6.x and a 7.x dashboard
Some issue where not all documents are indexed
If you're certain that docs are missing from all beats that points to 2, esp. since heartbeat doesn't have a dashboard.
Can you verify which beats you're actively using? @lucasfcosta and I are heartbeat experts, but if this is mostly a metricbeat issue you'll want to speak to a metricbeat expert.
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.