I want to monitor my spark jobs via ELK.
I am using the Graphite sink to send spark metrics to Logstash.
I want to define a Line Charts visualization in Kibana for jvm metrics. But the received fields names contain the spark application id (ex: application_1512022801087_025.driver.jvm.pools.Metaspace.committed)
Is it possible to have a generic field name to keep visualization works when the spark application id changes?