Hello Caroline,
First, the story copy issue is a bug of iceScrum R6#14.3, thanks for the report. We will probably release a fix version soon. I think that it has nothing to do with CPU usage.
If the CPU usage comes from a bug in the environment (the OS, Java or Tomcat itself) then there is not much we can do to help. Otherwise, if it comes from iceScrum then I can think of two possible causes:
-
Spawning of threads which are never killed an consume CPU (e.g. unsuccessful connections to the DB that won’t close). However, the graphs you have uploaded on imgur (this is the first time I use imgur for actual work, thanks for that!) indicate a brutal rising while continuous thread spawning would be more linear.
-
Heavy garbage collection, which happens when the JVM heap memory reaches the maximum defined by the -Xmx option. The JVM tries to free memory to avoid reaching the limit by removing dead objects but if there is a memory leak or if the Xmx is too low then it is doomed to fail with a « java.lang.OutOfMemoryError » error. This would typically cause a brutal rising of the CPU usage as the memory reaches its limit.
In order to be sure of the cause, can you please provide more information:
– JVM options (including Xmx and MaxPermSize): usually defined in CATALINA_OPTS
– Tomcat configuration file: server.xml
– Log files: icescrum.log and catalina.out
– Java version
– Tomcat version
You can write at support at kagilum dot com.
