Ethernet fails on IPCop / Linux

[Unchained]

Referring to http://www.mybroadband.co.za/vb/showthread.php?t=30943

I lost my connection through IPCop this evening, then couldn't connect again.
Even restarted the IPCop box, power cycled the UTD and restored last IPCop backup and then updated the UTD firmware eventually...
Then tried Fedora FC4, couldn't connect.
Then tried Windoze ethernet, no luck... Thought ethernet port on USB was blown...
Tried Win USB - fine.
Restarted Windoze, then ethernet was fine (and still is), but still no luck with either IPCop or FC4.

BTW, the direct telnet to UTD access is still fine, so also utTraceStar, but I can't seem to connect...

Any ideas on this issue?

 

[Bulldog]

Ethernet fails on IPCop

I'm glad i'm not the only one.....I've also changed the firmware on the modem back to the previous version, changed network cards, reinstalled ipcop a few times changed mtu settings....I tested the ethernet with windows xp, works no problem. I phone tech support, they say nothing has changed on their side and that they don't know anything about linux.....Please help!


and

 

[Bulldog]

Some more info

I enabled connection debugging, this is what I get :

00:44:47 red: Connectioncheck Restarting 5/5
00:44:47 red: Connectioncheck Clearing RED interface
00:44:50 pppd[1220] Plugin rp-pppoe.so loaded.
00:44:50 pppd[1220] RP-PPPoE plugin version 3.3 compiled against pppd 2.4.2
00:44:50 pppd[1221] pppd 2.4.2 started by root, uid 0
00:44:50 pppd[1221] PADS: Service-Name: ''
00:44:50 pppd[1221] PPP session is 10
00:44:50 pppd[1221] using channel 1
00:44:50 pppd[1221] Using interface ppp0
00:44:50 pppd[1221] Connect: ppp0 <--> eth1
00:44:50 pppd[1221] Couldn't increase MTU to 1500
00:44:50 pppd[1221] Couldn't increase MRU to 1500
00:44:50 pppd[1221] sent [LCP ConfReq id=0x1 <magic 0x1f732c19>]
00:44:51 pppd[1221] rcvd [proto=0xff03] c0 21 01 01 00 12 01 04 05 98 03 04 c0 23 05 06 f8 50 32 81 00 00 00 00 00 00 00 00 00 00 00 00 ...
00:44:51 pppd[1221] Discarded non-LCP packet when LCP not open
00:44:51 pppd[1221] rcvd [proto=0xff03] c0 21 02 01 00 0a 05 06 1f 73 2c 19 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ...
etc..........
...............
00:47:21 pppd[1221] LCP: timeout sending Config-Requests
00:47:42 red: Connectioncheck RED fail to connect
00:47:42 pppd[1221] Terminating on signal 15.
00:47:42 pppd[1221] Connection terminated.
00:47:42 pppd[1221] Exit.

any ideas?

 

[seburn]

Fix your mtu! Then see if there is still a prob?
http://www.unchained.co.za/ipcop.html

 

[Unchained]

Nope, not MTU / MRU prob... Even changed the pppoe call to use -m 1352
Log, most recent first:

08:11:04 red: Connectioncheck RED fail to connect
08:11:02 pppd[284] Exit.
08:11:02 pppd[284] Script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1352 -S iBurst finished (pid 286), status = 0x1
08:11:02 pppoe[286] Sent PADT
08:11:02 pppoe[286] read (asyncReadFromPPP): Session 15: Input/output error
08:11:02 pppd[284] script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1352 -S iBurs t, pid 286
08:11:02 pppd[284] Waiting for 1 child processes...
08:11:02 pppd[284] Connection terminated.
08:11:02 pppd[284] LCP: timeout sending Config-Requests
08:10:02 pppd[284] sent [LCP ConfReq id=0x1 <mru 1352> <magic 0x3c972513>]
08:09:02 pppd[284] sent [LCP ConfReq id=0x1 <mru 1352> <magic 0x3c972513>]
08:08:59 pppd[284] Discarded non-LCP packet when LCP not open
08:08:59 pppd[284] rcvd [proto=0xff] 03 c0 21 02 01 00 0e 01 04 05 48 05 06 3c 97 25 13
...
...
08:08:35 pppd[284] rcvd [proto=0xff] 03 c0 21 02 01 00 0e 01 04 05 48 05 06 3c 97 25 13
08:08:35 pppd[284] sent [LCP ConfReq id=0x1 <mru 1352> <magic 0x3c972513>]
08:08:34 pppd[284] Discarded non-LCP packet when LCP not open
08:08:34 pppd[284] rcvd [proto=0xff] 03 c0 21 01 02 00 12 01 04 05 98 03 04 c0 23 05 06 35 a9 d7 66
08:08:32 pppd[284] Discarded non-LCP packet when LCP not open
08:08:32 pppd[284] rcvd [proto=0xff] 03 c0 21 02 01 00 0e 01 04 05 48 05 06 3c 97 25 13
08:08:32 pppd[284] sent [LCP ConfReq id=0x1 <mru 1352> <magic 0x3c972513>]
08:08:31 pppd[284] Discarded non-LCP packet when LCP not open
08:08:31 pppd[284] rcvd [proto=0xff] 03 c0 21 01 01 00 12 01 04 05 98 03 04 c0 23 05 06 35 a9 d7 66
08:08:31 pppoe[286] PPP session is 15
08:08:31 pppoe[286] PADS: Service-Name: 'iBurst'
08:08:30 pppd[284] Connect: ppp0 <--> /dev/pts/0
08:08:30 pppd[284] Using interface ppp0
08:08:30 pppd[284] using channel 1
08:08:30 pppd[284] pppd 2.4.2 started by root, uid 0

 

[DaveBuchanan1337]

Number 1, get rid of IpCop - it's totally bloated,

m0n0wall is more secure, and is fully functional for a 4mb install.

Back to your question, have you tried reconfiguring your ethernet cards in shell -> setup?

Also, what cards are you using? 8132too?

 

[BadBoyZ]

Aaargh, the same (more or less) happened to me.
All of the sudden last night my connection just dropped and I couldn't get it up again. I'm running through an access point. Viewing the logs indicate that the access point manages to establish a link to the UT-D but then fails to connect to WBS.

If I connect the UT-D directly to my pc everything is fine. I didn't change a single setting and it was working fine for 3+ months. This has to be something on their side, I just don't know what. At first I thought maybe they banned routers, but then I set mine to clone a MAC (which would fool them into thinking it is a single pc), but still no luck.

Anybody got any ideas?
I'm running Windows, by the way.

 

[LandyMan]

08:11:02 pppd[284] script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1352 -S iBurs t, pid 286

I am by no means a linux expert ... but I see your -S flag is iBurs t ... can't this be a problem? don't know, just saw it

 

[8321]

Using bloated IPcop 1.4.10, no problems so far. No firmware upgrades.

 

[Unchained]

... but I see your -S flag is iBurs t ... can't this be a problem? don't know, just saw it

Must be due to copy & paste or IPCop log display, settings are correct, but thanks
Nov 15 18:34:13 ipcop pppd[2022]: script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1352 -S iBurst, pid 2023
Nov 15 18:34:13 ipcop pppd[2022]: Script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1352 -S iBurst finished (pid 2023), status = 0x1

Number 1, get rid of IpCop - it's totally bloated,

m0n0wall is more secure, and is fully functional for a 4mb install.

Back to your question, have you tried reconfiguring your ethernet cards in shell -> setup?

Also, what cards are you using? 8132too?

IPCop worked fine for me for the last 7 - 8 months or so, also on the exact same hardware. It happened suddenly yesterday without any changes. The reason I disconnected was to restart the UTD since (like before) I lost access to the UTD from my Windoze box. The net interfaces are fine, even utTraceStar accesses the UTD through IPCop and I can also telnet it from FC4 through IPCop, but can't connect from IPCop or FC4.

Aaargh, the same (more or less) happened to me.

Happened out of the blue, no response yet from WBS helldesk, Shaun Green or Luis.

Using bloated IPcop 1.4.10, no problems so far. No firmware upgrades.

Was my situation until it happened...

 

[Unchained]

IPCop reinstalled, still cannot connect.
Also still NOT on FC4, similar to IPCop errors (LCP: timeout sending Config-Requests)
Stuck to my pet hate: M$ Windoze, unless WBS fix what they've broken on Wonderboom or provide a solution

I have once again mailed Shaun Green, Luis and helldesk and also CC'd Sasan Parvin and referred them to this thread. Not going to hold my breath however...

 

[native]

Ok fine, now lets see
Logic tells me that iBurst have become MTU based connection technology.

Firstly, couple of users in Joburg started experincing problem with browsing but can authenticate on the radius server.

Secondly, those using routers with their MTU not set to 1352 / below had the same problems. [lazy to google the forum].

Thirdly, there are MTU's for 2000 pro on their site. 2000 doesnt have a build in PPPoE unlike Windoze XP. You have to be logged in to set the MTU.

Fourthly, most of the XP users on iBurst have been told to install DRTCP by helpdesk and now they dont know how to do it on IPCOP. [I will test it on smoothwall without the MTU set to 1352 and see if I will have sucess]


If I'm wrong then I'm wrong and WBS must come with a better explantion

 

[Unchained]

I'm not even going to bother further, have already notified them that I want to cancel.

Over the last few weeks there's been a couple of serious probs: Personal information leakage (Computing SA article), extended service interruptions (the most recent since last night till this evening in Pta), absolutely no customer service or communication from the start. Total ignoring of emails, etc, etc.

I just had enough...

Note: Although I've experimented with MTU settings on all the platforms I use (WinXP, FC4 & IPCop) and had to set it for WinXP to solve certain problems, I never had to set MTU in either FC4 or IPCop, although I tried it in an attempt to solve the current issue. Too large an MTU has certainly never blocked me from connecting / signing on...

 

[seburn]

I haven't heard anyone say that wbs read my mail ... In my opinion if you want help you have to phone the idiots at the call center ... They should be sensible to put you through to a more technical person if available

 

[Bulldog]

MTU issues

Hi there,

I managed to connect again by switching to monowall, quite cute so small, yet so effective. I have one question though, why does it not use the specified mtu in the config? In the logs it seems that it is switching to 1492 instead of the specified 1352. In any case it seems to be quite stable now.......touch wood. I'm also upset that WBS does not inform us on any changes made impacting mtu settings etc. I used to connect fine with Ipcop with the mtu on 1492.

Unchained, If you want a linux box to connect to Iburst, get Monowall, it works for me. I also went through 4 reinstalls and different network cards etc. before giving up.

I also noticed that we're surfing through wbs's proxy again, wonder when they will make up their mind......

Anyway, thanks for the tip on MonoWall!

 

[native]

I'm also upset that WBS does not inform us on any changes made impacting mtu settings etc. I used to connect fine with Ipcop with the mtu on 1492.

Communicating with clients is not part of their policy.
Getting answers from the call centre is like squeezing blood out of a stone

 

[Unchained]

Frustating and total waist of time

This WBS zoo is totally frustating and a waist of time...

The Wonderboom tower was down for a while again this morning and now both my IPCop and FC4 connect without a hint of a problem. After I went against my better judgement and even reinsalled IPCop, actually knowing it was a WBS problem and not on my side, as in numerous previous cases.

And not even a hint of a response or explanation from them But they can call at 08:00 out of the blue when it suits them to help identify a problem...

I basically waisted 3 days on a WBS problem! But once again the problem got "magically" resolved when the stirring started on MyADSL...

 

[seburn]

Typical

What problem? Was there a problem? We have no records of any emails or calls from you! You must be dreaming coz your the only one that thinks there is a problem.

 

[Unchained]

What problem? Was there a problem? We have no records of any emails or calls from you! You must be dreaming coz your the only one that thinks there is a problem.

Yeah, and you never get read receipt, even if digitally signing the message and requesting a secure receipt. I make a point of addressing my mails to various people along with helldesk, almost always signed and always requesting a read receipt and (in the mail) a response by email, not phone, sometimes even bcc'ing certain people on myadsl and often referring to a myadsl thread with the problem.

[Edit]
Sasan Parvin & Pierre de La Motte responded by email and requested that I provide a contact number, albeit only this evening and the issue was already resolved automagically...
[/Edit]

 

[8321]

Timeout waiting for PADO packets

I am still battling to connect a UTD with IPCop. The device works perfectly via Win 2000/usb but since last week it has not connected (having worked perfectly for more than a month).

I changed mtu and mru in /etc/rc.d/rc.red to 1352.
I updated the firmware - all WBS could suggest.
From the log below, there is still another mtu setting (1412??) that I don't know where to find in order to change it.

Nov 22 08:42:36 p219 ipcop: Dialling iBurst.
Nov 22 08:42:36 p219 pppd[10048]: pppd 2.4.2 started by root, uid 0
Nov 22 08:42:36 p219 pppd[10048]: using channel 13
Nov 22 08:42:36 p219 pppd[10048]: Using interface ppp0
Nov 22 08:42:36 p219 pppd[10048]: Connect: ppp0 <--> /dev/pts/0
Nov 22 08:42:37 p219 pppd[10048]: sent [LCP ConfReq id=0x1 <mru 1352> <magic 0xfe7252db>]
Nov 22 08:43:10 p219 last message repeated 11 times
Nov 22 08:43:11 p219 pppoe[10050]: Timeout waiting for PADO packets
Nov 22 08:43:11 p219 pppd[10048]: Script /usr/sbin/pppoe -p /var/run/pppoe.pid -I eth1 -T 80 -U -m 1412 finished (pid 10050), status = 0x1
Nov 22 08:43:11 p219 pppd[10048]: Modem hangup
Nov 22 08:43:11 p219 pppd[10048]: Connection terminated.
Nov 22 08:43:11 p219 pppd[10048]: Exit.

PS - I do not want to install m0n0wall, as a matter of principle.

I will be grateful for any ideas.