George,
I am running a much later version of diald (0.99.4) under Slackware 7.0
and recently had a very similar problem. Over a period of a couple of
weeks my connection was kept alive for very long periods of
time. Fortunately, I have 2 things in my favor. #1, I pay a flat rate per
month for my phone usage. That will change some day I'm sure; but not
yet. #2, I am good friends with one of the principals of my ISP. (Which
is the main reason they are my ISP. <smile>) I am NOT a Linux/network
guru. I am a generalist programmer with strong trouble-shooting
abilities. All things VMS are my specific bailiwick. Here's what little I
understand about what happened.
My ISP has expanded to support a wireless system which uses part of the
local "direct TV" system. Apparently they need to poll the wireless
modems. I did see an unusual message in my /var/log/messages file, for
example:
Aug 3 12:11:50 moe sendmail[5841]: NOQUEUE: Null connection from
blahblah.codenet.net [aaa.bbb.ccc.ddd]
I got hundreds if not thousands of these stamps in my messages file. This
may not have been the exact packet which kept me up but was part of the
whole problem and told my friend quite a bit. The ISP's tech who set up
this "polling" had their system polling ALL modems including the telephone
modems. That polling was apparently akin to, but not, a ping. (I believe
I've filtered out ping packets from keeping me up.) This polling was often
enough, every 2-5 minutes, to fake out my diald and convince it to keep an
otherwise idle link up.
I am sorry that I don't know the exact specifics of what was done by
the ISP. I do know, something else has now changed and my link now goes
down when I am idle for 5 minutes. If this sounds like it might be the
same problem that afflicted you, I can get greater, more professional,
detail from my ISP friend. He is their "part-time" system admin/network
guru and has complete technical knowledge of what goes on. His full-time
job is working at the same company where I work so talking with him isn't a
problem.
BTW - I too tried a reboot. (<sigh> My Linux system had been up
without a hiccup since 4-March-2000.) In my case it changed nothing about
the problem.
- Nelson ...
===================================================================
Nelson E. Ingersoll <> Desk: (719) 540-1263
Principal Software Engineer <> FAX: (719) 540-1006
ATMEL Corporation <> Pager: (719) 921-7917
1150 E. Cheyenne Mtn Blvd. <> INTERNET: [EMAIL PROTECTED]
Colorado Springs, CO 80906 <> [EMAIL PROTECTED]
United States of America <>
================== No good deed goes unpunished! ==================
At 8/9/00 06:37 AM, George Mailer wrote:
>I have been using diald-0.16.4-1 on a RedHat Linux version 2.0.32 system for
>several
>years without problem.
>
>I just came back after a day away to find it has been connected to my ISP
>for over 30
>hours ( at local telco rates in the UK !). ( It was running 'normal' and not
>'forced').
>
>- the normal rules didn't apply and upon investigation I found that sl1 was
>running - not sl0
> as normal - I can't see why ???
>
>- I tried restarting diald and it didn't clear the problem so for ease went
>for the panic button
> and did a reboot - all seems OK again now.
>
>- there would presumably have been no traffic on the link for the majority
>of this time.
>
>Does anyone have any ideas as to what could have been wrong and how might be
>best to
>prevent a re-occurance !
>
>Thanks in advance - regards - George.
>
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-diald" in
>the body of a message to [EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-diald" in
the body of a message to [EMAIL PROTECTED]
