[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Request for info

Chuck Myers wrote:
> 
> Yes, Steve gave me some ideas to check out, but in the end I changed the
> ldmadmin start script to include the log file location. I could get the ldm
> started manually, but only after a reboot. The verbose error messages on my
> receiver machine lead me to believe it was an RPC problem. But on closer
> inspection, the log file was being created as a zero byte file, and staying
> that way. I re-did the syslog.conf and that did not help. I kept messing
> with it and finally got a log to work at start up. I then modified the start
> script. I have not touched it since :) Its been running rock solid. This is
> on an HP270 machine. My receiver machine is an old HP915, it runs fine. Both
> run HP-UX 10.2
> 
> cm
> 
> Anne Wilson wrote:
> 
> > Hi Chuck,
> >
> > I was scanning through our archives and I came across some old email
> > from you regarding too many LDM processes running.  We also had a brief
> > discussion about run levels and when the portmapper and the LDM should
> > be invoked.
> >
> > This issue has popped up again for us.  Currently I'm suspecting network
> > unavailability.
> >
> > From your email it sounded like you and Steve Danz were going to
> > investigate your problem.  Were you ever able to identify what was
> > causing all the rpc.ldmd processes?
> >
> > Thanks,
> >
> > Anne
> > --
> > ***************************************************
> > Anne Wilson                     UCAR Unidata Program
> > address@hidden                  P.O. Box 3000
> >                                   Boulder, CO  80307
> > ----------------------------------------------------
> > Unidata WWW server       http://www.unidata.ucar.edu/
> > ****************************************************

Thanks, Chuck!

So, are you saying that once you got the logging working the problem of
too many rpc.ldmd processes went away?  If so, perhaps there's a bug in
the ldm such that if a sender or receiver process gets hung, e.g., in
trying to write to the log, the parent rpc.ldmd doesn't kill it, but
spawns another anyway.  

We'll be testing along these lines sometime in the future.

Anne
-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                 P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************