This letter is a bit out of topics, but somehow connected with ppp.
We have dial-in server on Linux box: Debian 2.0, kernel 2.0.34,
mgetty 1.1.14-2 (with default init string), pppd 2.3.5-2,
Apache Ae56SP-R external modem, serial speed 115200, hardware
flow control, answer after 3rd ring.
A few months all worked OK, but in last weeks we encontered veird
problem: sometimes (about once from a ten calls) modem don't
answer after third ring. Some minutes later all is OK. That is
only problem, no problems if connection is established (as rule
33600).
In case of normal calls mg_ttyS1.log (with debug 5) looks also
normal:
10/23 02:50:02 yS1 waiting for ``RING''
10/23 02:50:02 yS1 got: [0d][0a][0d][0a]RING ** found **
10/23 02:50:08 yS1 send: ATA[0d]
10/23 02:50:08 yS1 waiting for ``CONNECT''
10/23 02:50:08 yS1 got: [0d][0a]ATA[0d][0d][0a]CONNECT ** found **
In case of failed call mg_ttyS1.log looks like:
10/23 02:27:43 yS1 send: ATA[0d]
10/23 02:27:43 yS1 waiting for ``CONNECT''
10/23 02:27:43 yS1 got: [0d][0a]ATA[0d][e6][80][98][80][18]f[18]
[86][98]~[98][e6][80][98][80][e6][80][98][80][18]f[18][86][98]~[98]
[e6][80][98][80][e6][80][98][80][18]f[18][86][98]~[98][e6][80][98]
[80][e6][80][98][80][18]f[18][86][98]~[98][e6][80][98][80][e6][80]
[98][80][18]f[18][86][98]~[98][e6][80][98][80][e6][80][98][80][18]
f[18][86][98]~[98][e6][80][98][80][e6][80][98][80][18]f[18][86][98]
~[98][e6][80][98][80][e6][80][98][80][18]f[18][86][98]~[98][e6][80]
[98][80][e6][80][98][80][18]f[18][86][98]~[98][e6][80][98][80][e6]
[80][98][80][18]f[18][86][98]~[98][e6][80][98][80]
10/23 02:29:03 yS1 timeout in chat script, waiting for `CONNECT'
10/23 02:29:03 ##### failed timeout dev=ttyS1, pid=5408,
caller='none', conn='', name=''
It looks like communication problem between modem and Linux box.
Mgetty don't get CONNECT and after timeout resets modem, then next
call is OK.
Any clue?
I suspect, that right in time, when Linux box send ATA, modem
received sometimes next RING (between ring of sending side and
receiving side is significant delay). Then probably will help
replacing of ATA to ATH1 (wait for OK) AT~~~A, but it needs
recompiling of mgetty. Also here is few users with summary about 10-
20 dialling in a day, so collecting of some statistics needs time.
Maybe somebogy can understand something of this garbage (instead of
CONNECT). Every time it's just same (sometimes longer, sometimes
shorter, but you can see, there is repeated same sequence).
Best regards,
Valentin Abramov
E-Mail: [EMAIL PROTECTED]
WEB: http://www.tartu.customs.ee/valentin.shtml
-
To unsubscribe from this list: send the line "unsubscribe linux-ppp" in
the body of a message to [EMAIL PROTECTED]
