[H-GEN] ppp pleez..

[Robin & Andy McPherson-Bruce]

Craig Eldershaw wrote:

> Who can't you talk to ?  Can you not reach any computer other than
> your
> own ?

'rampo' (5x86, ne2k clone, WinGate) plays gateway for 'andrax' (5x86,
ne2k clone, RH4.2) - from rampo, posting to dejanews just left me
hanging endlessly waiting for a 'reply from host', hence my inability to
communicate with 'anyone'. my isp's group server hasn't functioned
(usefully) since early August (quote : "Subject: News server. Date: Wed,
12 Aug 1998 13:05:36 +1000   From: (name deleted to protect the
innocent(?) <him_again at nrg.com.au>
     To: rampo at nrg.com.au
"Yes I know about the news server and I'm looking into it. Strange thing
is it works fine from here (NRG home office) it's only the dial-up users
that are having problems. So I've already been onto Telstra (who run the
news server) and we're trying to resolve it." hmm.).

> Several diagnostic tests:
>
> ifconfig
>
>     Should say something comforting about ppp0 being up.

hmmm.. it prob'ly does, briefly - tho' i'd have to be quick. fyi, from
debugging -

Sep 20 19:01:51 andrax pppd[4602]: pppd 2.3.4 started by root, uid 0
Sep 20 19:02:18 andrax pppd[4602]: Serial connection established.
<snip>ppp_tty_ioctl flags<snip>
Sep 20 19:02:19 andrax kernel: ppp: write frame, count = 24
Sep 20 19:02:19 andrax kernel: FF 03 C0 21 01 01 00 14 ...!....
Sep 20 19:02:19 andrax kernel: 02 06 00 00 00 00 05 06 ........
Sep 20 19:02:19 andrax kernel: 3F 8A C8 69 07 02 08 02 ?..i....
Sep 20 19:02:19 andrax pppd[4602]: Using interface ppp0
Sep 20 19:02:19 andrax pppd[4602]: Connect: ppp0 <--> /dev/cua1
Sep 20 19:02:19 andrax pppd[4602]: sent [LCP ConfReq id=0x1 <asyncmap
0x0> <magic 0x3f8ac869> <pcomp> <accomp>]
<snip>multiple ppp: write frames<snip>
Sep 20 19:02:49 andrax kernel: ppp: channel ppp0 closing.
Sep 20 19:02:49 andrax kernel: ppp0 released
Sep 20 19:02:49 andrax kernel: ppp0: ccp closed
Sep 20 19:02:46 andrax last message repeated 9 times
Sep 20 19:02:49 andrax pppd[4602]: LCP: timeout sending Config-Requests
Sep 20 19:02:49 andrax pppd[4602]: Connection terminated.
Sep 20 19:02:49 andrax pppd[4602]: Receive serial link is not 8-bit
clean:
Sep 20 19:02:49 andrax pppd[4602]: Problem: all had bit 7 set to 0
Sep 20 19:02:50 andrax pppd[4602]: Exit.


> ping <your-own-ip>
>
>     If this doesn't work, then networking is badly broken at your end
> -
>     bring it into Humbug for a checkover. :-) (oh, and ip, means a
>     number [eg.  11.222.123.221], not a hostname).  If you have a
>     dynamicly allocated ip, then skip over this for a moment.

we're dynamically allocated; pap-secrets has an entry for
my_username_at_nrg, i've also tried including an empty user/secret entry
with the remote ip address (no luck) & 'noauth, noipdefault' options to
pppd (ditto).

> ping <ppp-other-end>
>
>     Again, you want a number here, not a name.  This is the ip of the
>     machine at your isp's end.  You probably have it as an option to
> ppp
>     on the right hand side of a `:'.  If this doesn't work then the
> ppp
>     link itself is at fault.

no time & no 'real' connection - i get 9 write frames & LCP times out,
almost invariably - plus the 'not 8-bit clean'. if i dial in via
minicom, immediately after sending my p'word i get :'PPP session from
(203.16.176.2) to 203.16.176.*** beginning...
pppd: peer authentication required but no authentication files
accessible
NO FSCKING CARRIER'

no sh*t..

> ping 130.102.2.20
> ping student.uq.edu.au
>
>     If the first works, but not the second, then you have a DNS
> problem,
>     check your /etc/host.conf, /etc/resolv.conf and /etc/nsswitch.conf
>
>     files.  If they both work, then waht are you complaining about ?
> :-)
>     If neither works, then your routing is probably broken; check your
>
>     default route (give the command "route", or if that fails, try
>     "route -n").
>
> If still stuck, then report back to the list with the results of the
> above tests, and a copy of your ppp configuration file/command-line
> options.

fr. /etc/ppp/ppp-on : exec /usr/sbin/pppd -d /dev/cua138400 kdebug 7
CONNECT $dialler script(have tried /dev/ttyS1, /dev/modem with identical
results; & yes, it _is_ the modem port 8~) )

fr. /etc/ppp/options : -detach defaultroute modem crtscts asyncmap 0
lock noipdefault user rampo noauth

fr. /etc/host.conf : order hosts, bind \n multi on

fr. /etc/resolv.conf : domain nrg.com.au \n nameserver 203.16.176.2

fr. /etc/nsswitch.conf : passwd: files nisplus nis \n shadow: (as above)
\n group: (as above)

(hadn't come across /etc/nsswitch.conf - have i missed something?)

> Hope this helps.
>
> Cheers,
>         Craig.

given the continuing sorry state of nrg's news server, and given that
this is the same sysadmin who told me i _could not_ login via a script
(i _must_ use ezppp/tkppp (& m. anthon is correct; it _is_ just a pretty
front end for the _exact_ same scripts)), and given that i have now
observed, identified & rectified the lack of synchronicity between
development strands of ppp.c (kernel/pppd), and given that pppd sets the
designated serial port to 8-bit clean by default anyway, i'm beginning
to think that maybe i'm not the problem..

but hey, assumptions can kill..

love & strawberries

andy b.