In message <Pine.BSF.email@example.com>, Tom writes:
>On Thu, 20 Jul 2000, Jared Chenkin wrote:
>> In message <200007202035.e6KKZ9V14881@dna.tsolab.org>, "Dan Ts'o" writes:
>> >> Does su have some kind of a built-in nohup option? If I su to root and
>> >> execute a command or shell script and then disconnect (ie, quit the terminal
>> >> software I'm running, which in my case is an ssh session) whatever I was
>> >> last running su'd as root continues to run until I manually kill it.
>> Random question, but do you actually log out or do you just close the
>> ssh window? I notice that very often users on my system simply close
>> their telnet windows and the process does not die (namely the shell).
>> It became a real annoyance when telnetd(8) would start turning away
>> successful logins, complaining that all ttys were used up.
> Good Telnet clients should actually close the socket before quiting, and
>good OSes will automatically close any sockets left open when an
>application quits. Either of these would eliminate the problem you see.
>However, if a client crashes TCP keepalives should eventually timeout the
>socket on the server. There is a sysctl that forces TCP keepalives for
>all sockets that can be handy for services that don't necessarily use
>them. However, the telnetd service on FreeBSD uses keepalives by default.
We're talking about the telnet client that comes with Windows 95/98 ... I
didn't say that there was anything *good* about it :)
Thanks for the feedback but I was possibly attempting to make a point to
an earlier query (if you didn't notice..) The problem no longer persists
on my system, but thanks for the thought.
Networked Systems Administrator
Bronx Science Computing
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message