On Wed, Oct 22, 2008 at 03:27:10PM -0700, Wayne Cochran wrote: > We have a Linux file server for a set of computer science and > engineering > labs where each lab contains machines running Mac OS X, Ubuntu Linux, > or some incarnation of Windows. At times the these machines become > almost unusable and I think I have narrowed the problem to smbd > processes > soaking up all the CPU on the server. Running 'top' on the server > reveals a dozen or so entries like the following: > > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND > 18229 millerbr 20 0 12420 3980 2992 R 26 0.1 0:03.85 smbd > 18225 tande8 20 0 12380 3564 2660 S 22 0.1 1:00.19 smbd > 18861 dhenniga 20 0 12676 4316 3328 R 7 0.1 0:01.86 smbd > 18210 samcprui 20 0 12616 4064 3128 S 3 0.1 0:55.84 smbd > 18214 mlhorn 20 0 12568 3912 3060 S 3 0.1 0:25.23 smbd > 18867 tsze 20 0 12784 4656 3592 S 3 0.1 0:10.28 smbd > 18728 daflores 20 0 12684 4560 3516 S 2 0.1 0:35.44 smbd > 18678 twharris 20 0 12780 4556 3552 S 1 0.1 0:11.35 smbd > ... > > Looking through the log files reveals nothing obvious at first glance, > but of course > I am not sure what to look for and there are a lot of log files. > > What steps should I first take to debug the situation?
What version of Samba are you running on this server ? What is the operating system on the server ? What does tracing the smbd execution with strace, or ptrace say when smbd is spinning like this ? Jeremy. -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
