regards
Venugopal
OUR BEST SOLUTION SO FAR:
We insist they go back to login first prior to closing windows
or we start locking offenders out until their supervisor has trained
them properly.
This does not say which OpenSSH you've installed. I know that 5.0.6,
for example, had a reasonable package at
ftp://ftp.sco.com/pub/openserver5/opensrc/openssh-4.2p1/openssh42p1_vol.tar.
But note that there's nothing that will force the new sessions to the
same terminal session for the same users. Is it clear they've
actually closed the session? Is there a shell still running associated
with it? Closing the window *should* kill the connection on the client
end, which should end the forked off daemon on the server end and
gracefully clear the session.
The next user who logs in to *that pseudo-terminal*, yes, exactly.
Irrelevant of what /etc/hosts says, what does "ifconfig" say about
your system's network devices? Does it perhaps have one listed at
192.168.100.1, perhaps for the second of several network ports? You've
gotten me curious about why it had that spurious entry. A lot of
people do very, very strange things to /etc/hosts.
Sometimes the utmp(x) and wtmp(x) are not syncronised. Try to empty them
in single user mode and check again.
John
No, that isn't normal.
Are the users using a terminal emulator on a PC to connect to the
server? If so which emulator? Do they use telnet or SSH as the
communications protocol?
What command are you using to find out who is logged-in? Are you using
the 'who' command?
--
RGB
I've noticed the behaviour using TinyTerm to telnet into 5.0.5
systems. Logging out of applications may, or may not, clear assignment
to the psuedo terminals. In my cases login is not running and the
terminal is reassigned the next time it is needed. Whodo shows the
terminals but no process ids or jobs assigned to them.
Look into 'cwtmp'. See: http://www.sco.com/ta/105755
--
JP
Note that Anzio Lite and AnzioWin (our terminal emulation programs for
Windows) can be configured to prevent simply quitting. This coerces the
user into logging out of the server app's menu system. If course nothing
can prevent them from rebooting the PC.
We have also seen sites where someone has had a 'trap' in a shell script,
which effectively prevented the hangup signal from reaching the running
application.
More information at
http://www.anzio.com/support/kb/Anzio_kb/What%20happens%20when%20a%20user%20quits%20a%20Telnet%20session.htm
Regards,
....Bob Rasmussen, President, Rasmussen Software, Inc.
personal e-mail: r...@anzio.com
company e-mail: r...@anzio.com
voice: (US) 503-624-0360 (9:00-6:00 Pacific Time)
fax: (US) 503-624-0760
web: http://www.anzio.com
street address: Rasmussen Software, Inc.
10240 SW Nimbus, Suite L9
Portland, OR 97223 USA
They are using putty (telnet) to login.
> What command are you using to find out who is logged-in? Are you using
> the 'who' command?
>
w command is used to check logged in users.
regards
Venugopal
Oh, dear. First thing to do then is to not use telnet, and upgrade to
using OpenSSH. Among other benefits, such as its security benefits,
I'll pretty much guarantee you that it better handles disconnected
sessions than ancient telnet servers. Plus, it handles X sessions in
far more graceful fashion: this makes it very useful for remotely
connecting to a SCO server and running graphic administrative tools,
like scoadmin.
netconfig only adds new entries to /etc/hosts.
That means if you change your ip in netconfig, or install new nics, or
delete & reinstall the same or a different nic, etc etc in all cases it
just adds a new line to /etc/hosts, so it's common to have old incorrect
entires in there unless you know about this and fix it manually after
every use of netconfig. Similarly the broadcast address in
/etc/default/tcp may bear any or no relation to the current ip settings.
--
bkw