WAN port dies - fixing speed

[Oleg]

We need some info on this terrible problem, so Antiloop as well as number of other guys have reported, that fixing port speed fixes the problem.
So, the idea is to prove this finally. :)

Anyway, please try to fix port speed to 100full, 100half, 10full or 10half. Try to find one, which fix your problems.

Please note, that fixing port speed will slow down boot-up for the first time. This is due to the fact, that partner on the other side of the link will need to determine port speed and connection type by itself.

Also, it's possible, that link will never be established (possible reason for that is dumb device on the other side). Wait several minutes before claiming "it does not work".

Another option, pereviously reported on the forum is that wl500g does not boot with forced speed - just disconnect WAN port cable and power off/on unit. It's safe.

[Styno]

Never happened to me.

Modem brand and type: Zyxel prestige 650R-31 (Analog ADSL modem/router).

[wtzm]

Anyway, please try to fix port speed to 100full, 100half, 10full or 10half.
Do mean something like
wlanutil eth1 ethsetx speed 4 (100full, in this case)?

Try to find one, which fix your problems.
I have tried all possible speed settings and the wan port still dies after some time of heavy traffic. Interestingly, the transfer speed I experience isn't influenced by the speed I set with wlanutil.

[Oleg]

Do mean something like
wlanutil eth1 ethsetx speed 4 (100full, in this case)?

I mean changing it from web interface or with et utility by exusting something like this from the shell prompt, or Main_AdmStatus_Content.asp:


et -i eth1 speed 100full

Please re-test, it should have impact on performance. :)

[wtzm]

I mean changing it from web interface
Well, this setting isn't available to me as my wl500g's "Operation Mode" is set to "Router" (my wan port is connected to another segment of my home network, not to an adsl modem or anything like that...)
Therefore I took a look at /init/ethernet

if [ -n "$x_ConnectionSpeed" ] && [ "$x_ConnectionSpeed" != "0" ]; then
wlanutil ethsetx $1 speed "$x_ConnectionSpeed"
fi

and thought setting the Connection-Speed is done with wlanutil.

I'll post my results when setting the speed with the et utility after some tests.

[wtzm]

Setting the speed to 10half, 10full, 100half or 100full with the et utility,
there were no crashes even after transferring a few GBytes of data.
Using the auto setting eth1 dies (and can be recovered with et -i eth1
promisc 0) after some hundred MBytes.

Test setup:
[pc2]-eth0 <~~~> eth1-[wl500g]-eth0 <~~~> eth0-[pc1]
pc2 uses a Realtek 8139C for the connection to the wl500g,
test traffic was created with netio-1.17

[Oleg]

Setting the speed to 10half, 10full, 100half or 100full with the et utility,
there were no crashes even after transferring a few GBytes of data.
Using the auto setting eth1 dies (and can be recovered with et -i eth1
promisc 0) after some hundred MBytes.

Test setup:
[pc2]-eth0 <~~~> eth1-[wl500g]-eth0 <~~~> eth0-[pc1]
pc2 uses a Realtek 8139C for the connection to the wl500g,
test traffic was created with netio-1.17
Well, Antiloop reported the same - any forced speed fixes problem for him. Added new poll option - any speed fix. :)

[smallkeung]

For me, WAN Port die issue only happen on downloading with BT. When the transfer rate reach 8xx KB, it MUST die within 5 mins. (My internet connection is 10mb/10mb VDSL) Actually I'm ASUS fan, but this product really makes me DISAPPOINTTED. :mad: :mad: :mad:

[Antiloop]

For me, WAN Port die issue only happen on downloading with BT. When the transfer rate reach 8xx KB, it MUST die within 5 mins. (My internet connection is 10mb/10mb VDSL) Actually I'm ASUS fan, but this product really makes me DISAPPOINTTED. :mad: :mad: :mad:

can you please confirm that you have tried all speeds and not only AUTO ?

e.g. when i'm testing i have 2,5Megabyte/sec download and then it does not die when using a fixed speed setting

try firmware 1.8.1.9, in that firmware the speed settings work from webinterface in earlier 1.7.x.x versions it does not work or not correctly

[smallkeung]

can you please confirm that you have tried all speeds and not only AUTO ?

e.g. when i'm testing i have 2,5Megabyte/sec download and then it does not die when using a fixed speed setting

try firmware 1.8.1.9, in that firmware the speed settings work from webinterface in earlier 1.7.x.x versions it does not work or not correctly

i tried to use 1.8.1.9. but the problem still occurr. So, i changed to use 1.7.5.9 cr-5 with the QOS. actually, it work if i limited in a slow speed. But i don't think that it's not a good way to fix the wan port die problem.

By the way, i'm very very appericate the working in web site. I'll try to
upgrade it back to 1.8.1.9 and try all port speed.

THANK YOU VERY MUCH

[Thorz]

Styno says: OOPS, sorry, I edited your post instead of quoting it. I'm terribly sorry. Please post it gain ...

:o

[Antiloop]

@ jcbos
can you please confirm that you have tried it correctly (e.g. with new 1.8.x.x firmware) to test if line dies when using a fixed speed ?

[Thorz]

Styno says: OOPS, sorry, I edited your post instead of quoting it. I'm terribly sorry. Please post it gain ...

:o

Man, It was a long one. I don't remember all that was put there. It is a pity because there was some good info there for helping to diagnose this problem. If I remember later I will try to post it again.

[GJAman]

Here are the results from tests I preformed about two and a half months ago:

After I entered in telnet “et -i eth1 speed ....” (where the dots have to be replaced with auto or full10 or half10 or full100 or half100. I transferred up to 9 big files (+/- 700 MB from WAN to LAN using ASUS M3700N as FTP server on WAN side and a PC as FTP client on the LAN side)



et -i eth1 speed auto
RESULT: WAN dies in a few minutes

et -i eth1 speed full10
RESULT: after 20 minutes the WAN is still alive

et -i eth1 speed half10
RESULT: after 20 minutes the WAN is still alive

et -i eth1 speed full100
RESULT: doesn't work at all (WAN light doesn't even go on)

et -i eth1 speed half100
RESULT: doesn't work at all (WAN light doesn't even go on)


BTW: I have a 1120/352 Kbit/s internet connection connected with a SpeedTouch 510 an the WAN side

[MSG]

Hi, I've already tried setting the speed of my router but it still seems to drop off every now and then when I use BT. After it disconnects (PPPoE dropping off), I would have to do a manual disconnect then a connect and then do the et -i eth1 promisc 0 command to get the PPPoE connection back up again. I'm very annoyed at this and it would be greatly appreciated if a fix can be formulated for this.

I am currently using the 1.8.1.7-1 firmware.

I don't know if the following log would be useful or not, but these seems to be generated when my PPPoE connection dropped off.


Oct 14 14:46:03 (none) local2.notice pppd[2821]: Serial link appears to be disconnected.
Oct 14 14:46:03 (none) local2.warn pppd[2821]: Couldn't increase MTU to 1500
Oct 14 14:46:03 (none) local2.err pppd[2821]: Couldn't increase MRU to 1500
Oct 14 14:46:03 (none) user.emerg PPPoE: Disconnected
Oct 14 14:46:09 (none) local2.notice pppd[2821]: Connection terminated.
Oct 14 14:46:09 (none) local2.warn pppd[2821]: Doing disconnect
Oct 14 14:46:39 (none) local2.err pppd[2821]: invalid packet
Oct 14 14:46:39 (none) local2.err pppd[2821]: Failed to negotiate PPPoE connection: 9 Bad file descriptor


Also, I don't know if this is related, but sometimes, I get this line:

Oct 14 12:58:27 (none) user.warn klogd: ip_conntrack: table full, dropping packet.

can someone tell me what it is?

Thanks in advance

[Styno]

Also, I don't know if this is related, but sometimes, I get this line:

Oct 14 12:58:27 (none) user.warn klogd: ip_conntrack: table full, dropping packet.

can someone tell me what it is?


Yes, you've reached the maximum number of concurrent IP connections over the router. New connections are being refused -> dropped. This mostly happens when you're using P2P software. By default the router can handle 1024 concurrent connections. You can set this to a higher number like 2048 or 4096. This has been described in this forum before.

[MSG]

Yes, you've reached the maximum number of concurrent IP connections over the router. New connections are being refused -> dropped. This mostly happens when you're using P2P software. By default the router can handle 1024 concurrent connections. You can set this to a higher number like 2048 or 4096. This has been described in this forum before.
Firstly, Thanks for the reply~~=)

In my case, I've set my p2p client to accept no more than 80 connections at a time for each transfer concerned, and it shouldn't have more than 200 connections at a certain time since I don't do multi transfers often, then how come I manage to get more than 1024 concurrent connections? I'm quite puzzled about this. Also, could this be related to why my PPPoE connection drops off??

I've done a search in how to set my concurrent connection to a higher number, but since I'm now using the new firmware of 1.8.1.7-1, is it still the same? and do I just execute them in the web interface?

nvram set bootCmdXXX="echo 4096 > /proc/sys/net/ipv4/ip_conntrack_max"
nvram commit

Thanks

[Thorz]

Firstly, Thanks for the reply~~=)

In my case, I've set my p2p client to accept no more than 80 connections at a time for each transfer concerned, and it shouldn't have more than 200 connections at a certain time since I don't do multi transfers often, then how come I manage to get more than 1024 concurrent connections? I'm quite puzzled about this. Also, could this be related to why my PPPoE connection drops off??


Are you using Azureus? If so I will post my example for you:

I am NOT using PPPoE, just a plain LAN ADSL 2048/256 connnection
I have set in Cinfiguration / Transfer the following:

Azureus ver. 2.1.0.4 official.
Max # of connections per torrent: 80
Max # of connections globally: 500
Default max uploads slots per torerent: 3

When doing multidownloads of popular torrents I have checked in Netlimiter 1.30 (excelent not only for bandwidth shaping but for the statistics that it gives you, recommended) and I have found that active connections number made from Azureus can go higher than the max global # of 500. It has been sometimes as high as 800, but never reaching the 1024 limit. This has been downloading from a known and popular torrent site and usually having 5-6 active torrents (2 seeding, 4 downloading). I recommend you checking the real actual connection # with any program that gives you this info, remember that it is not only Azureus using your connection at a given time, there are other apps doing it also. If you find that it is not Azureus then I recommend you checking against spyware or other nasties if you have not yet done so.

[Antiloop]

Firstly, Thanks for the reply~~=)

In my case, I've set my p2p client to accept no more than 80 connections at a time for each transfer concerned, and it shouldn't have more than 200 connections at a certain time since I don't do multi transfers often, then how come I manage to get more than 1024 concurrent connections? I'm quite puzzled about this. Also, could this be related to why my PPPoE connection drops off??

I've done a search in how to set my concurrent connection to a higher number, but since I'm now using the new firmware of 1.8.1.7-1, is it still the same? and do I just execute them in the web interface?

nvram set bootCmdXXX="echo 4096 > /proc/sys/net/ipv4/ip_conntrack_max"
nvram commit

Thanks

FYI bootCmdX does NOT work anymore in custom 1.8.x.x related firmwares it's obsolote, for more info: read the thread where you got the custom firmware from.

[smallkeung]

i've tested the 1.8.1.x for a month. (1.8.1.9 and 1.8.1.7-2a)

The disconnect problem is sitll occurred when i was using Bit Torrent. I'm not sure that what will cause it happen. Sometimes it can run it for a day. Sometimes it runs couple mins then it will disconnect. I guess the problem may caused by when too many connection and a higher transfer occurr at the same time then the connection will be disconnected

The following are the Log after disconnect from firmware 1.8.1.7-2a

Jan 1 09:29:27 smallkeung-nas user.warn klogd: Adpt 0 c000
Jan 1 09:29:28 smallkeung-nas user.warn klogd: Link up c000 8100
Jan 1 09:29:28 smallkeung-nas user.warn klogd: Adpt ffffffff 8100
Jan 1 14:25:44 smallkeung-nas user.warn klogd: Link down c033
Jan 1 14:26:07 smallkeung-nas local2.notice pppd[87]: Serial link appears to be disconnected.
Jan 1 14:26:07 smallkeung-nas local2.warn pppd[87]: Couldn't increase MTU to 1500
Jan 1 14:26:07 smallkeung-nas local2.err pppd[87]: Couldn't increase MRU to 1500
Jan 1 14:26:08 smallkeung-nas user.emerg PPPoE: Disconnected
Jan 1 14:26:13 smallkeung-nas local2.notice pppd[87]: Connection terminated.
Jan 1 14:26:13 smallkeung-nas local2.warn pppd[87]: Doing disconnect
Jan 1 14:26:20 smallkeung-nas user.warn klogd: Adpt 2 c000
Jan 1 14:26:21 smallkeung-nas user.warn klogd: Adpt 1 c000
Jan 1 14:26:22 smallkeung-nas user.warn klogd: Adpt 0 c000
Jan 1 14:26:23 smallkeung-nas user.warn klogd: Link up c000 8100
Jan 1 14:26:23 smallkeung-nas user.warn klogd: Adpt ffffffff 8100
Jan 1 14:26:43 smallkeung-nas local2.err pppd[87]: invalid packet
Jan 1 14:26:43 smallkeung-nas local2.err pppd[87]: Failed to negotiate PPPoE connection: 9 Bad file descriptor
Jan 1 14:26:43 smallkeung-nas daemon.err pppoe-relay[82]: PADO packet from 00:e0:fc:10:29:48 on interface eth1 does not have Relay-Session-Id tag


Oct 28 21:21:38 smallkeung-nas user.emerg ntp client: time is synchronized to asia.pool.ntp.org
Oct 28 22:53:09 smallkeung-nas user.warn klogd: Link down c033
Oct 28 22:53:13 smallkeung-nas local2.notice pppd[87]: Serial link appears to be disconnected.
Oct 28 22:53:13 smallkeung-nas local2.warn pppd[87]: Couldn't increase MTU to 1500
Oct 28 22:53:13 smallkeung-nas local2.err pppd[87]: Couldn't increase MRU to 1500
Oct 28 22:53:14 smallkeung-nas user.emerg PPPoE: Disconnected
Oct 28 22:53:19 smallkeung-nas local2.notice pppd[87]: Connection terminated.
Oct 28 22:53:19 smallkeung-nas local2.warn pppd[87]: Doing disconnect
Oct 28 22:53:20 smallkeung-nas user.warn klogd: Adpt 2 c000
Oct 28 22:53:21 smallkeung-nas user.warn klogd: Adpt 1 c000
Oct 28 22:53:22 smallkeung-nas user.warn klogd: Adpt 0 c000
Oct 28 22:53:23 smallkeung-nas user.warn klogd: Link up c000 8100
Oct 28 22:53:23 smallkeung-nas user.warn klogd: Adpt ffffffff 8100
Oct 28 22:53:49 smallkeung-nas local2.err pppd[87]: invalid packet
Oct 28 22:53:49 smallkeung-nas local2.err pppd[87]: Failed to negotiate PPPoE connection: 9 Bad file descriptor
Oct 28 22:53:49 smallkeung-nas daemon.err pppoe-relay[82]: PADO packet from 00:e0:fc:12:f4:ca on interface eth1 does not have Relay-Session-Id tag
Oct 28 22:53:49 smallkeung-nas daemon.err pppoe-relay[82]: PADO packet from 00:e0:fc:0f:39:d8 on interface eth1 does not have Relay-Session-Id tag
Oct 28 22:53:49 smallkeung-nas daemon.err pppoe-relay[82]: PADO packet from 00:e0:fc:10:29:48 on interface eth1 does not have Relay-Session-Id tag

I've a stupid question. What's the meaning for the WAN port die are you guys talking? At the beginning, i upgraded to 1.8.1.x from 1.7.1.x The Router can't connect to the internet. Then i discovered that the WAN port light didn't on. Finally, I changed the Port from auto to 10mb Full. This problem is fixed. I never see the WAN port light off.
If this kind of problem is the WAN port die issue of you guys talking, then my problem is another disconnect problem.

[MSG]

FYI bootCmdX does NOT work anymore in custom 1.8.x.x related firmwares it's obsolote, for more info: read the thread where you got the custom firmware from.

I can't seem to work out how I can set the ip_conntrack value to a higher value, can someone please teach me the steps I should take (for new firmware 1.8.1.7-1) to set it higher?

Also, can someone tell me the command I should use to disconnect and connect when I use telnet or through the web manager interface?

Thanks

[Styno]

I can't seem to work out how I can set the ip_conntrack value to a higher value, can someone please teach me the steps I should take (for new firmware 1.8.1.7-1) to set it higher?

Have you tried the webinterface? If it does not work, you can also try to figure out what that options does in the source of the page and try to debug it.

[Oleg]

I can't seem to work out how I can set the ip_conntrack value to a higher value, can someone please teach me the steps I should take (for new firmware 1.8.1.7-1) to set it higher?

Also, can someone tell me the command I should use to disconnect and connect when I use telnet or through the web manager interface?

Thanks
1.8.1.7-2a has "maximum connections to track in the web iface"

[Oleg]

I can't seem to work out how I can set the ip_conntrack value to a higher value, can someone please teach me the steps I should take (for new firmware 1.8.1.7-1) to set it higher?

Also, can someone tell me the command I should use to disconnect and connect when I use telnet or through the web manager interface?

Thanks
1.8.1.7-2a has "maximum connections to track in the web iface"

[kiboko]

Hi everyone!!

My wan port is still not working properly. I've tried fws 1.8.1.9, 1.8.2.4 and 9.8.2.1 and wan is going down.

I've also tried all the port speeds and the results were'nt good.

speed auto
RESULT: Takes a lot of time to connect(WAN dies in a few hours)
speed full10
RESULT: doesn't work at all (WAN light doesn't even go on)
speed half10
RESULT: doesn't work at all (WAN light doesn't even go on)
speed full100
RESULT: Takes a lot of time to connect(WAN dies in a few hours)
et -i eth1 speed half100
RESULT: Takes a lot of time to connect(WAN dies in a few hours)

My modem is a dlink dcm-200, could it be modem's problem????


I don´t know what to do!!! :confused: :mad:

[DaSuperGrover]

Now running FW 1.8.2.4, and WAN port doesn't die anymore :cool:

My WL-500G is in AP mode, and now connection to first router (also Linux :D ) can stay in the WAN port :)

[kjetil]

Hello.

Setting my Asus 500g to 100MBps fixed speed the WAN die problem has totaly disappeared. This was a major problem for me bacause I have a 10mbps/10mbps internet connection, and you can imagine how few bytes you can transfer before the WAN died.

edit: used with success with 1.8.1.7-1, -2a and -3

[xtraya]

I have 5 units in my wifi network, and EVERY 500g have problem when speed reach more than 2 MB/s

especialy if you connect two 500G, one from LAN to another WAN port.
Just for few second after booting, WAN LED just turns off and on non stop if there is any kind of more speedy transfer. If it reach 2 MB/s or more, port dies.

I tryed to fix speed on all speeds, but it wont work because it just stops to work within 1 or 2 seconds.

I tryed every useable firmware revisions, but everything is the same.
I found 1.8.2.4 BETA on asus site, where they wrote that they fixed WAN problem, but after uplaoding, WAN is completly DOWN...

I saw that if ASUS and unit on the other WAN side are connected over SWITCH (10/100 with shorter cables) it dies but after 2-3 hours of continous 3-4 MB/s transfer

This is a problem, big one... other bugs are cosmetic-like

[Styno]

Have you tried resetting to defaults after firmware upgrade? This might solve your problem...

[xtraya]

yes, I tryed everything, with and without reset, with different UTP cables,models of switch,with other units (linksys wrtg54g,pheenet 118G,dlink dwl2000+)

and result is the same,if transfer reach 2,5 MB/s and more, dies within a 10-20 sec

[Oleg]

did you use dhcp for address assignment?

[dr-snuggles]

I wanted 2 thank u all at the forum. First i suffered from a general connection loss, but after installing the 1.8.2.4. firmwar that was a little solved. I had my azureus on upnp and like 5-6 times a day a mesages connection lost. Also, my surfing days were gone whenever i was downloading with azureus.

According 2 advis here in the forum i shut off the upnp and put my modem 2 100full duplex and.... fulle donwload speeds and all computers on my network (3 wired clients, 2 wireless clients) still able to surf, games etc. Wonderful :D

[dr-snuggles]

Maybe i cheered a little 2 soon. After 5 days it choked again :mad: . Th estrange thing is my torrents keep running, just the rest isn´t accessible anymore :confused:

[Oleg]

This is a different problem. Increase max number of tracked connections to 4096

[virrr]

I had I dieing wan port aswell.

I upgraded the firmware to v1.8.2.4 BETA and set speed to fixed 100 full and I had not had a disconnnect since then.

Just for your information.

[dr-snuggles]

Oleg, thanx for the tip. I have read the thread about it, but couldn't find that setting at my firewall settings. And someone else said that it was fixed in the 1.8.XXX firmwares??? Or i am very dumm, but i can't find the connections setting anywhere (firmware 1.8.2.4)

[Oleg]

well, you need custom firmware for this... 1.8.1.7-3 is latest.

[Oleg]

Alternatively you can write down a script (file), which could be executed from usb drive if you've one, the contents should be



#!/bin/sh
echo 4096 > /proc/sys/net/ipv4/ip_conntrack_max

[dr-snuggles]

Oleg, thanx again for the support. I will try 2 put the custom firmware in since i don't have a usb stick or any. My scripting knowledge is zero so i am afraid i will f**k up the router if i try that.

[freestuff]

Problem happened to me also. :mad:
I have a Zyxel prestige 650R-31 (Analog ADSL modem/router), the problem only happens when I'm connected with my laptop (laptop with a Intel 2100 wireless card) I'm using the Asus firmware 1.8.1.9 and I'm using it as a home gateway.
Most off the times it happens when I'm downloading from Usenet (with the laptop) at full speed 250-300KB. Sometimes the Asus wan port dies after 1 hour.
I will try the workaround (fixed port speed).
I will post the results after a few days.

[xytron]

I hva prettu much the same problem as dr-snuggles and I'm also using Azureuz. But the problem is that i don't know how to change the port speeds?!?! So can someone please tell me exactly how to put my modem to 100full duplex?!?
I've also upgraded to firmware 1.8.2.4.
Hope you could help me out here. I hate the "wan-die" problem!!!

[ronak]

The general rule is: Same settings in both ends ! This means that if you ain't able to configure the other end, leave to auto to let autonegotiation work correctly. With 100Mb be aware to use kables of GOOD qality, min. CATegory5.

[jago25_98]

I haven't been able to get the WAN port working at all.

The settings are simple:

ifconfig eth1 192.168.100.10 on the router

and

ifconfig eth0 192.168.100.11 on the computer

using a crossover cable that has been tested and various media types / auto-negoitation.

- cannot ping 192.168.100.10 from the computer
- WAN light is lit up
- cannot ping 192.168.100.11 frin the router but this is probably because the ping command on the router probably doesn't support the -I switch

How can I be 100% sure my WAN port is broken because I personally know the guy who sold it to me. Could I send it back? Could I somehow add an ethernet port, replace the WAN port or replace the wireless card for an ethernet card?

[tomilius]

First word of advice is not to assume something is broken just because you don't know how to use it. Second word of advice is to try the forum search. Third--have you even tried using the WAN port the way it's supposed to be used?

[Styno]

jago, I just read a post from you where you state that the WAN port works in AP mode, so I guess it's a configuration problem...

[akira]

hello, after months of cursing and trying out all of the above things, I found out something what fixed it for me.

I thought that the router was a little hot and placed it on it's side,before I laid flat all of the time.

that was it for me :) , I hope it also works for other Routers, please post something if it does. (just curious :confused: )



mzzl AkIrA

[tomilius]

I thought that the router was a little hot and placed it on it's side,before I laid flat all of the time.
That's pathetic... not the solution, but that so many wireless routers seem to have this problem! Does anyone even THINK of overheating during production? Sheesh.

[ela002]

Any solution about the WAN port? I use PPPoe on ASUS to connect via USR 9105 ADSL modem while I ping the LAN IP of the ASUS. When there is a little activity on WAN port (internet) I get ping time out. How can we fix that? It happens to all latest firmwares customs or officials.


:(

[Antiloop]

since firmware 1.8.2.4 BETA this thread has become obsolete

hereby I'm closing the thread as the problem has been solved by asus


Please use firmware version 1.8.2.4 or newer at your option.