[1038] in linux-net channel archive
Re: Questions for ppp services
daemon@ATHENA.MIT.EDU (Steven K.A. Tkaczuk)
Mon Sep 4 07:52:03 1995
Date: Sun, 3 Sep 1995 22:18:53 -0400 (EDT)
From: "Steven K.A. Tkaczuk" <steve@procom.procom.net>
To: hyoon <hyoon@hitel.kol.co.kr>
cc: linux-net@vger.rutgers.edu
In-Reply-To: <Pine.SOL.3.91.950902105708.29632B-100000@hitel.kol.net>
On Sat, 2 Sep 1995, hyoon wrote:
> Hello everyone,
>
> Our system environments are as follows:
> CPU: pentium 75
> Main Memory: 32MB
>
> Kernel version: 1.3.3
> slackware: 2.2
> pppd version: 2.1.2d
> ppp channels: 48
>
> We opened the PPP services three days ago.
> I have got three questions.
>
> Q1: I want to know the correct number of who is currently logged on,
> and be forbidden when a user who is previously logged on try to log on.
>
> I confirmed that ther messages of 'who' is not correct by using 'ps'.
> The users of ppp services do not have shell prompt,
> but directly execute 'pppd'.
>
> /etc/passwd:
> soo0613:HpO2DeFxoTJbk:1002:300:soo0613::/usr/etc/ppplogin
> ...
>
> Q2: Error Messages in /var/adm/syslog are as follows:
>
> Aug 31 16:53:25 mingkey pppd[577]: ioctl(PPPIOCSASYNCMAP): I/O error
> Aug 31 16:53:25 mingkey pppd[577]: ioctl(PPPIOCGDEBUG): I/O error
> Aug 31 16:53:25 mingkey pppd[577]: ioctl(TIOCSETD): I/O error
> Aug 31 16:53:25 mingkey pppd[577]: ioctl(TIOCNXCL): I/O error
> Aug 31 16:53:25 mingkey pppd[577]: tcsetattr: I/O errorB
>
> What do these mean?
>
> Q3: The system is frequently downed.
> But there are not the messages in /var/adm/messages, etc.
> We are very serious.
>
>
> Thanks in advance.
I run an ISP using PPP and have encountered the same problems you have..
Hope these answers help..
Well, in answer to question 1.. If I understand it you just want to
stop/block multiple logins? If so, there is a program called loginhog that
limits the user to only one open session at a time.
Also we are running a login script that does the dynamic PPP, but still I
have no problems tracking users with a who..
As for question 2.. The error seems to come from when your caller/user
hangs up. The pppd seems to wait for about 3-4 seconds then
realizes something is wrong. I encounter the same problem and although
it fills the messages log with those error codes I have found no ill
effects. I have experimented with killing the PID for the pppd and if done
while the user is still online, the errors are not generated, I then tried
to add something similar to the pppd login script so it would kill the
pppd on termination of the script instead of teh 3-4 second hang. I sorta
have a functional way of removing the errors, but there are a few glitches
as well. I have seen about 3 other people put up messages about the same
type of errors and if anyone has a conclusive answer, please let me know..
If you want to know what I have tried so far.. mail me. :)
Answer for 3, did you look in teh /var/adm/syslog to see about messages?
I get the occasion system crash but it is always OOPS pid 1: init()
which then locks up my users from exiting any sessions, it is most annoying
but awaiting a tape backup before attempting to tackle it.. It does seem
to generate that OOPS when a user does hang up though, so maybe it is
related to the pppd (but it only happens every 10-12 days or so..)
Any suggestions would be appreciated..
Steve Tkaczuk
(steve@mail.procom.net)
--Ergo Eudora Sum... I E-mail, therefore I am.
