[torqueusers] Torque 2.1.x pbs_server process hogging cpu

garrick at speculation.org garrick at speculation.org
Tue Jun 13 01:42:06 MDT 2006


On Tue, Jun 13, 2006 at 09:19:40AM +0200, Martin Schafföner alleged:
> On Monday 12 June 2006 21:40, garrick at speculation.org wrote:
> > On Mon, Jun 12, 2006 at 04:22:09PM +0200, Martin Schafföner alleged:
> > > Today I felt like doing some updates, so I first tried upgrading from
> > > torque 2.0.0p7 (not too old, I guess) to torque 2.1.0p0. Installing the
> > > software went fine; however, when I now submit a job, the job isn't
> > > executed. Instead, the pbs_server process eats all of the available CPU
> > > time.
> >
> > Can you get some process traces or gdb backtraces and see where it is
> > looping?
> 
> So, w.r.t. torque-2.1.1-snap.200606072052, pbs_server is looping in 
> src/lib/Libnet/net_client.c in function client_to_svr(...), in the "retry" 
> loop between lines 185 and 307. I will poke around a little to see if I can 
> catch the obvious...

First determine the state of HAVE_BINDRESVPORT and NOPRIVPORTS in
src/include/pbs_config.h.

What OS is this?

It looks like that goto needs to be wrapped around an #ifndef
HAVE_BINDRESVPORT.  But it should still exit that function if tryport
gets too low.



More information about the torqueusers mailing list