my data exists in daily indices and 10-20 milion doc per indice ,with 32G memory for elastic jvm .
when i run a dashboard elastice run at least 20 query that created by visualizes .
, when i set time range between 1 and 7 days, All things are good.
but for more days range, i have multiple issue:
1_ kibana query timed out after 3 minute (seted by elasticsearch.requestTimeout).
2_ elastic Continues to execute query for awhile after that 3 minutes.
3_ breake down status appers.
4_ i added more memory and no benefit.
there is no problem if i set some filters (reduce the number of docs ) .
i need some setting to restrict executing concurrent query count that created by dashboard, eg: 5 cuncurrent from all visualize in dashboard.
and wait bar for another loading visualizes .
i see long new generation gc and i think these steps occeurred:
1: while new generation gc , elastic is fully stopped
2: kibana timeout occurred
3: finally query execute continuing until finish
in my dashboard exists visualizes like line,bar,gauge,pie,region map ,data table,timelion and other.
i finally found that "data table" take high amount of ram and it is cuse of breake down.
my data table is top 10 source ip s that have more distinct destinations :
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.