On Fri, 15 Feb 2002, $Bill Luebkert wrote: > Carl Jolley wrote: > > > On Wed, 13 Feb 2002, $Bill Luebkert wrote: > > > >>Sounds like you're running out of file descriptors (usually 64 or 256 > >>are available without extending). My guess is that Perl is not > >>closing/re-using the descriptors when you re-assign them after the > >>spawn. > >> > >> > > > > Should not each iteration of the while loop (in which $stderr and $stdin > > are defined as my variables) result in perl closing the file and > > releasing the file descriptor? The DESTROY method of the super class > > (IO::Handle) is specficially designed to do nothing but perl should still > > do its thing on the deactivation of the objects. > > > Should is the appropriate word. :) > > > > If not wouldn't: > > > > $stdout->close; > > $stderr->close; > > > > just prior to (or after) the reopening of STDOUT and STDERR do the > > trick? > > Tried that when testing it too. :) >
And what effect did calling the close method on the objects have? I assume that based on your lack of a confirming statement that doing so did not clear up the problem. Did your testing show why or based on your testing can you speculate as to why closing the underlying file handles did not free up the file descriptors? Did your testing confirm or support your guess that the problem was a failure of perl to release and re-claim file descriptors? Does this appear to be a general problem with perl's global cleanup/memory deallocation or does it seem to specifically relate to file descriptors? **** [EMAIL PROTECTED] <Carl Jolley> **** All opinions are my own and not necessarily those of my employer **** _______________________________________________ Perl-Win32-Users mailing list [EMAIL PROTECTED] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
