On Mon, Dec 20, 2004 at 03:10:09PM +0100, Andrea Gangini wrote: > > Hi, I noticed this behavior on one of my server (mail server, postfix + > amavisd + spamassassin, 2,8ghz Xeon, kernel 2.6.9-1.681_FC3smp). > > Here's the output of top command: > > top - 15:04:35 up 4 days, 6:33, 2 users, load average: 15.00, 15.03, ..... > Any clues/suggestions? ..... I have seen this same symptom recently. As best I could tell some cron started tasks blocked on a lock file access involving a LONG running task (setiathome). Since I had done a number of critical path updates and system admin stuff that could have placed these tasks in harms way I rebooted and have not seen the problem since. Kevin posted a hint to use "ps" with a long alphabet soup of flags. When looking at that output also look at "lsof" output and match it to the processes (a network resource could be the problem). Since you are running a multi processing (SMP) kernel it may be that you have a set of processes or threads that are in a true live lock situation. It is likely that a user space process is spinning on some application specific lock (resource contention). If so then the application needs to be debugged. Listings of ps and lsof can help narrow the list of processes for debugging. -- T o m M i t c h e l l spam unwanted email. SPAM, good eats, and a trademark of Hormel Foods.
