New users can't login to one of my server . The cause is that freed ttys
are not re-used for new connections. Some users use Exceed.
Adding more ttys is a good way to solve this problem..
Any suggestions would be very much appreciated.
There isn't a question here.Quote:> New users can't login to one of my server . The cause is that freed ttys
> are not re-used for new connections. Some users use Exceed.
> Adding more ttys is a good way to solve this problem..
If you're trying to release the pty's, look for processes. If, for
example, /dev/pts/12 were in use from an old session, do this..
/usr/ucb/ps -t pts/12
That will report all processes associated with that pty. Killing them
(or finding out why the're not dying automatically) should return the
port to use.
If you're interested in increasing the number ptys available to users,
see the Solaris FAQ.
http://www.science.uva.nl/pub/solaris/solaris2.html#q3.41
--
Unix System Administrator Taos - The SysAdmin Company
Got some Dr Pepper? San Francisco, CA bay area
< Please move on, ...nothing to see here, please disperse >
>There isn't a question here.
>If you're trying to release the pty's, look for processes. If, for
>example, /dev/pts/12 were in use from an old session, do this..
>/usr/ucb/ps -t pts/12
It will report all processes with that pty as controlling terminal. It'sQuote:>That will report all processes associated with that pty.
/usr/sbin/fuser /dev/pts/12
Chris Thompson
Email: cet1 [at] cam.ac.uk
Excellent idea. Normally when dealing with 'stuck' ptys, I can findQuote:>>That will report all processes associated with that pty.
> It will report all processes with that pty as controlling terminal. It's
> possible to have a pty open without it being the controlling terminal,
> and this will also prevent the pty from being reused. To find such
> processes. use
> /usr/sbin/fuser /dev/pts/12
--
Unix System Administrator Taos - The SysAdmin Company
Got some Dr Pepper? San Francisco, CA bay area
< Please move on, ...nothing to see here, please disperse >
1. idle running of command which lost tty
I found many commands will idle running (and eat about
80% CPU time) when these commands originally should
run in a tty but lost tty now. I found these situation
many times and commands include tin and elvis.
Is it the bug of these commands ? or something wrong
in our system. Users lost tty connections can be
caused by many reasons, such as inadequately quit
of window, disconnection of network suddently,...
And if connections were closed, these commands should
received SIGHUP signal and quit. But they don't quit
and instead idle running in the background.
I doubt these commands use nonblocking read or
select calls to detect input, and not handle SIGHUP
cleanly. When there are no tty to be read, what will
be happened ? just idle running ?
--
Shih-Kun Huang | Huang2 Shih4-Kun1
2. Can someone direct me to an overview of shared libraries?
4. kdn
5. Linux 2.0.xx shutdown: lost tty output / hangs
6. run metadb against previously configured state database slices
10. tty hog - characters lost ?
11. Losing /dev/tty in a process
12. How to map a tty port to another tty port
13. creating tty's, mknod, stuck after boot with wrong tty - help