On Tue, Jul 01, 2003 at 04:50:22PM +0200, Frank Lichtenheld wrote: > On Fri, Jun 27, 2003 at 02:46:19PM +0200, Josip Rodin wrote: > > On Sun, Jun 22, 2003 at 12:06:13PM +0200, Frank Lichtenheld wrote: > > > - The script needs much memory (200 to 300 MB for the > > > unstable distribution) and is slow (but my computer has only 256 MB > > > RAM, I can not really speak for swapless execution) > > > Should I try to optimize it mainly for speed or for memory? > > > > How long does it take to run? > > During the last two days I tweaked it somewhat. It now needs ten > minutes on my Athlon 900 (512 MB RAM) for 'stable'. But it needs > up to 400 - 500 MB RAM during the run. (ca. 40-50% of the execution > time goes into whatever perl makes after my last command, I assume > it's calling desctructors and garbage collecting. If anyone knows a > trick to free the memory more "rude", I will happy to hear it ;)
Perhaps make sure you've flushed all I/O handles and then POSIX::_exit(0)? -- Colin Watson [EMAIL PROTECTED]
