PPP 2.2.0f connect to Xylogics Annex 4000 hangs

PPP 2.2.0f connect to Xylogics Annex 4000 hangs

Post by Roy C. Bixl » Sun, 18 Aug 1996 04:00:00



Hi,

I've posted before on this to 'comp.os.linux.networking', but have yet
to get any suggestions.  Any help would really be appreciated, since
I've spent a fair amount of time on this and tried quite a few tests
to no avail.  I can get my connection to completely freeze up just by
sending the linux-2.0.12.tar.gz file.

My system's a generic '486/40 with 16550A UART, 28.8 Motorola Power
modem and running Linux 2.0.13 with PPPD 2.2.0f.  My ISP uses a
Xylogics Annex 4000 and they have told me they have it set up to use
hardware flow control.  I doubt that this is a flow control problem
since I can send the file successfully using a terminal program and
Z-modem.  I suspect some interaction between PPPD and the Annex, since
I can also successfully send the Linux file with another PPP account I
have on a Gator box.  My PPP configuration is below:

38400
crtscts
modem
defaultroute
asyncmap 0
escape 11,13,ff
lcp-echo-interval 60
lcp-echo-failure 2
#debug
#kdebug 7

On one run, I also had the debugging options shown above enabled.
Here is the debug output just before the connection hung:

Aug 15 21:31:32 bix kernel: A7 8E 86 D1 49 10 39 D4 ....I.9.
Aug 15 21:31:32 bix kernel: 3D 10 87 6F 4B D1 79 11 =..oK.y.
Aug 15 21:31:32 bix kernel: 36 DF A0 DF E5 A8 76 49 6.....vI
Aug 15 21:31:32 bix kernel: 5F 45 C3 C0 E8 1E 68 06 _E....h.
Aug 15 21:31:32 bix kernel: 96 B4 65 C3 3C 81 41 1B ..e.<.A.
Aug 15 21:31:32 bix kernel: 0E BF 6D D1 A3 F9 31 BC ..m...1.
Aug 15 21:31:32 bix kernel: 08 4C D7 D1 70 88 F2 18 .L..p...
Aug 15 21:31:32 bix kernel: FA 36 B6                .6.
Aug 15 21:31:32 bix kernel: ppp_dev_xmit_lower: fcs is 2529
Aug 15 21:31:32 bix kernel: ppp_dev_xmit: writing 1481 chars
Aug 15 21:31:32 bix kernel: ppp_dev_xmit [ppp0]: skb 00ec9668
Aug 15 21:31:37 bix kernel: ppp_dev_xmit [ppp0]: skb 00ec9668
Aug 15 21:31:40 bix kernel: ppp_tty_read: called buf=08013702 nr=1504
Aug 15 21:31:40 bix kernel: ppp_tty_read: no data (EAGAIN)
Aug 15 21:31:46 bix kernel: ppp_dev_xmit [ppp0]: skb 00466284
Aug 15 21:31:50 bix last message repeated 5 times
Aug 15 21:31:55 bix kernel: ppp_dev_xmit [ppp0]: skb 00466888
Aug 15 21:31:58 bix last message repeated 3 times
Aug 15 21:32:01 bix kernel: ppp_dev_xmit [ppp0]: skb 00ec9668
Aug 15 21:32:02 bix kernel: ppp_dev_xmit [ppp0]: skb 00466888
Aug 15 21:32:10 bix kernel: ppp_dev_xmit [ppp0]: skb 00466888
Aug 15 21:32:24 bix pppd[26189]: sent [LCP TermReq id=0x2]

The last line represents my sending the 'pppd' process a HUP signal
after verifying the connection was indeed hung.  And here's from the
informational messages:

Aug 15 21:31:21 bix kernel: 6E 9C 17 27 BF 1C ..n..'..
Aug 15 21:31:21 bix kernel: 7>AB CA 9B 31 4E 79 5F 10 ...1Ny_.
Aug 15 21:31:24 bix kernel:  72 08 08 82 10 8F =.r.....
Aug 15 21:31:26 bix kernel: B 43 9C A2 63 C2 7F ..C..c..
Aug 15 21:31:28 bix kernel: 8F 09 3D 38 F9 7E F1 m..=8.~.
Aug 15 21:31:29 bix kernel:  0B 2A 15 07 7C CA FC Q.*..|..
Aug 15 21:31:30 bix kernel:  F5 CF 4D B4 A4 ,.(..M..
Aug 15 21:31:40 bix kernel: ppp_tty_ioctl: read demand dial info
Aug 15 21:32:24 bix kernel: ppp: channel ppp0 going down for IP
packets!

Lastly, my ISP tells me that the latest version of Trumpet works
without a hitch on the Annex and that they have some Linux boxes with
PPP connections that hang on occasion as well.  Anyone have any ideas?

Thanks!

Roy Bixler

--
"It's a damn poor mind that can only think of one way to spell a word."
                -- Andrew Jackson

 
 
 

PPP 2.2.0f connect to Xylogics Annex 4000 hangs

Post by Roy C. Bixl » Wed, 21 Aug 1996 04:00:00



>  I can get my connection to completely freeze up just by
>sending the linux-2.0.12.tar.gz file.

<various configuration and debug info deleted>

Quote:>Lastly, my ISP tells me that the latest version of Trumpet works
>without a hitch on the Annex and that they have some Linux boxes with
>PPP connections that hang on occasion as well.  Anyone have any ideas?

I have a bit more information now and thought I would post it since I
haven't gotten any useful suggestions (beyond "use 'tcpdump'", that
is) yet.  I was logged in remotely once when I got my PPP connection
to hang and I then logged into the Annex, ran a 'stats' command and
saw the Annex thinks I'm still connected, but the 'rts' signal is
low.  Also, running yet another test from the Linux end, the 'ftp' of
the Linux file stopped as usual but I could tell from the 'ping' to
the gateway I was running on another console that the connection was
still (barely) alive.  There was no major activity on the modem save
for the 'ping' packets and I noticed the round-trip time was over
1,000 ms, much higher than the normal 140-200 ms range.  Shortly after
that, the connection froze up completely.  Has anyone seen this
before?  Any help greatly appreciated!

Thanks,
Roy Bixler

http://bix.suba.com/~rcb/
--
"It's a damn poor mind that can only think of one way to spell a word."
                -- Andrew Jackson

 
 
 

PPP 2.2.0f connect to Xylogics Annex 4000 hangs

Post by Roy C. Bixl » Wed, 21 Aug 1996 04:00:00


I wrote (more or less, anyway):

Quote:>> I can get my connection to completely freeze up just by
>>sending the linux-2.0.13.tar.gz file.

>>My system's a generic '486/40 with 16550A UART, 28.8 Motorola Power
>>modem and running Linux 2.0.14 with PPPD 2.2.0f.  My ISP uses a
>>Xylogics Annex 4000 and they have told me they have it set up to use
>>hardware flow control.  I doubt that this is a flow control problem
>>since I can send the file successfully using a terminal program and
>>Z-modem.  I suspect some interaction between PPPD and the Annex, since
>>I can also successfully send the Linux file with another PPP account I
>>have on a Gator box.  My PPP configuration is below:

>>38400
>>crtscts
>>modem
>>defaultroute
>>asyncmap 0
>>lcp-echo-interval 60
>>lcp-echo-failure 2

OK, a couple of things have changed, but the result's the same.  I'm
now on kernel 2.0.14 and I removed my 'character escape' option.
Also, if anyone's interested, I have the output of a 'tcpdump' session
at:

http://www.veryComputer.com/~bix/pkt_trace.txt

For some mysterious reason, this time I got almost half-way through
the file (i.e. 3. Meg.) before the connection hung, so this is a bit
longish (350k).  (And, you guessed it, I don't put it on
'bix.suba.com' because I don't want people* my connection
trying to help me. :-) )

Help anyone?

Thanks again ...

Roy Bixler

http://www.veryComputer.com/~rcb/
--
"It's a damn poor mind that can only think of one way to spell a word."
                -- Andrew Jackson

 
 
 

PPP 2.2.0f connect to Xylogics Annex 4000 hangs

Post by Michael Tan » Thu, 22 Aug 1996 04:00:00




> >  I can get my connection to completely freeze up just by
> >sending the linux-2.0.12.tar.gz file.

> <various configuration and debug info deleted>

> >Lastly, my ISP tells me that the latest version of Trumpet works
> >without a hitch on the Annex and that they have some Linux boxes with
> >PPP connections that hang on occasion as well.  Anyone have any ideas?

> I have a bit more information now and thought I would post it since I
> haven't gotten any useful suggestions (beyond "use 'tcpdump'", that
> is) yet.  I was logged in remotely once when I got my PPP connection
> to hang and I then logged into the Annex, ran a 'stats' command and
> saw the Annex thinks I'm still connected, but the 'rts' signal is
> low.  Also, running yet another test from the Linux end, the 'ftp' of
> the Linux file stopped as usual but I could tell from the 'ping' to
> the gateway I was running on another console that the connection was
> still (barely) alive.  There was no major activity on the modem save
> for the 'ping' packets and I noticed the round-trip time was over
> 1,000 ms, much higher than the normal 140-200 ms range.  Shortly after
> that, the connection froze up completely.  Has anyone seen this
> before?  Any help greatly appreciated!

> Thanks,
> Roy Bixler

> http://bix.suba.com/~rcb/
> --
> "It's a damn poor mind that can only think of one way to spell a word."
>                 -- Andrew JacksonI have the same problem when connecting to Annex, it hang. But it works OK when

connects to my internet provider.  I am still trying to resolve the problem.
 
 
 

1. RH 3.0.3, Kernel 2.0, ppp 2.2.0f hang up

Hi All,
I would like to comment an extrange effect I have
observed in my laptop Compaq Aero.
I have 20MB RAM memory.
When I did installed Red Hat 3.0.3 and kernel 2.0
I did noticed that the 20MB RAM was not fully
recognize by the kernel so I did put a line to
loadlin at boot time. "mem=20M". With that line
the kernel will recognize the 20MB RAM but now
cames the problem:
always I put such line at boot time and I do
a PPP or SLIP connection after a while my laptop
is fully hang up (I have to switch it off and on to
continuing using it).
I will appreciate if anyone have any similar experience
or if someone could tell me to which person I should
address this problem.


thanks,

--
--------------oOo--------------------
Javi


SPAIN

2. How vunerable am I?

3. PPP not escaping properly on an Xylogics Annex

4. WHAT VIDEO CARD for X (VLBUS)???

5. Dialling-out via a Xylogics 4000 (and other questions)

6. Samba server stopped working

7. pppd 2.2.0f hangs my system

8. distribution Vs implementation

9. PPP Upgrade from 2.1.2d to 2.2.0f

10. ppp (2.2.0f) with kernel 2.0.11/13: OK not received.

11. ppp 2.2.0f and mschap80

12. PPP 2.2.0f compilation failure

13. ppp 2.2.0f carry ipx. How?