On 14/02/2011, at 10:09 PM, Benjamin Knoth wrote: > If i use archiva dump on 100 cpu usage it creates a dump of the java > process with a high pid. This process runs normally with 0 % cpu and 4.9 > % Mem. But if i restart archiva and make a dump with archiva dump i > didn't found a second java process on top.
This sounds a little weird. You have two java processes running? Or only when the dump is done? (haven't noticed that before - but the dump runs too fast to check). You can confirm what is expected by checking the PID in ./logs/archiva.pid. Is that the one on 100% CPU? > What should i do with the dump of archiva? Does the list of threads give any hint about what is going on in the server at that time? Perhaps you could attach that text to a jira issue for further investigation. We'd mainly be interested in anything that contains a trace from "org.apache.*" > > At the same time in meantime there are using max. 5 people archiva at > the same time. Given what you've shown in the logs, status page, and number of users so far, there should be no reason for the CPU to go to 100% at all, other than at pre-defined scanning intervals. I think we'll need to gather more logs or data via JIRA to figure out what's happening... - Brett -- Brett Porter br...@apache.org http://brettporter.wordpress.com/ http://au.linkedin.com/in/brettporter