The script connects to each portmaster in a list in turn, and gathers
"Who" records, and stuffs information about ports that need to be reset
onto a stack. It then starts popping items off the stack, reconnecting
to the portmasters, and killing off abusive users.
However(!):
It seems that it works fine for hours and hours, and then my mailbox
starts filling up with messages from cron saying that it couldn't
connect to one of the PMs. I check my logfiles and find that it is
getting the "Who" data from each of them, then supposedly disconecting
from the last one it looked at and then immediately reconnecting to it
(this is stack, remember!). When it tries to reconnect to the PM it
just disconnected from (with "close(portmaster'DS)"), it generates a
message saying that the connection was refused.
I saw a post here a few weeks back about resetting TCP port 1643 on
the PMs to take care of the problem, but I lost the message and have
been unable to recall how to reset the "Hnd" object shown by "show
netconns".
Anyone have any ideas?
Oh, and I'll post the script if anyone wants it...
Peace,
Mike
-- Mike Gogulski syadasti@iag.net Assistant System Administrator support@iag.net Internet Access Group Altamonte Springs, Florida, USA +1-407-786-1145 Work +1-407-672-2340 Other