Does anyone have any decent write up how setting JVM limits at Dispatcher and Query Service levels, affects report execution? Has anyone been able to figure out why quite often there are multiple java processes running consuming all available memory?
???
Hi,
in the same boat here. We have a 128GB RAM server with 8 CPU Cores and I am wondering what to put under:
JVM heap size limit query service (MB) (Currently I have 120GB)
Initial JVM heap size limit for the query service (MB) (25GB)
Is that too much, too little (I know minimum initial must be 4GB), but all the info on sizing is 5+ years old.
How about number of low and affinity connections and number of processes?