Darrin Barr wrote:

> Bill, Appreciate the quick response.  I thought about Win32::Process but ran
> into a different issue.
> 
> * Since I don't know how long the parent is going to be processing (anywhere
> from 3 minutes to several hours), the child can't be allowed to time out.
> I'll kill everything manually if the parent seems to have hung...

So don't time it out then.

> * I am trying to limit the # of processes on the system (part of what's
> being monitored), so starting another one for the "child" is less desirable
> than using fork

fork is kludged on Win32 - use at your own risk.

> * I need to let the child know when the parent's processing is done so it
> can write out some summary statistics to a file.  Killing the child prevents
> that.  That's why I was using a signal to communicate with the child.  If I
> use Win32::Process, the signal never makes it to the child (not sure why,
> yet).

There is no signal mechanism on Win32 - some of the logic is also kludged
to allow something similar.  My guess is that with threads, they have a
better signal emulation due to the fact that it's the same process.

You'll get more uniform processing without the fork, but signaling will
be a problem - better to poll from the child after each sleep and check
a semaphore or ?.

-- 
  ,-/-  __      _  _         $Bill Luebkert    Mailto:[EMAIL PROTECTED]
 (_/   /  )    // //       DBE Collectibles    Mailto:[EMAIL PROTECTED]
  / ) /--<  o // //      Castle of Medieval Myth & Magic http://www.todbe.com/
-/-' /___/_<_</_</_    http://dbecoll.tripod.com/ (My Perl/Lakers stuff)
_______________________________________________
Perl-Win32-Admin mailing list
[EMAIL PROTECTED]
To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs

Reply via email to