Hello, I have a server with dedicated cpu cores, so my server actually doesn't have a system cpu usage, but the spark plugin still shows an incorrect (because the value shouldn't exist) value of the cpu usage of the system

spark just reports what the JVM knows via the management interface.