Thanks for taking the time to reply.
No luck with the dev release, seems to same to me, I'm afraid. I should be able to knock up a little script that'll show the behaviour some time this weekend.
Incidentally in the script I got from CPAN it said 1.56_01 at the top even though the link was labelled 1.56_02.
| [reply] |
| [reply] |
Hi - Ok I have looked at this some more and the problem appeared to be only on one machine, and it sorted itself out after a reboot.
Interestingly (perhaps) before the reboot all of the plink pids in Windows' processes list were 16000 to 17999 or so (actually they may have been 10 X that, 160000 to 179999, not sure), while the pid that your module was trying to kill was always around 4000 or so. After the reboot the plink pids in the list are more normal, also around 4000. It wasn't just plink doing this before - all of the pids, bar one, of the processes started by the user I was logged in as and running stuff as were very high.
I did at one point wonder whether the machine having 8 cores could be a factor, but binding plink to a single core made no difference.
Many thanks for your help. After testing it on my home machine I could see it all working perfectly so I made sure everything (Perl versions etc.) was the same, tested again, found it still worked, then tried rebooting the server as a last resort. And it all seems fine now.
I'm guessing and hoping that this will just be one of those strange problems that will now disappear.
| [reply] |