Is this a bug on Asus firmware? WAN disconnections

[RMerlin]

Cant see that kind of item?



Checked all WAN tabs without an luck

Nevermind then. The issue you described usually happens to people with PPPoE or PPTP connections.

In your case, it was simply that your router had failed to obtain a lease from your ISP yet, so it was normal that 0.0.0.0 would be shown on the status page.

 

[rassi]

Nevermind then. The issue you described usually happens to people with PPPoE or PPTP connections.

In your case, it was simply that your router had failed to obtain a lease from your ISP yet, so it was normal that 0.0.0.0 would be shown on the status page.

Ok. Factory reset it is then.

 

[rassi]

In your case, it was simply that your router had failed to obtain a lease from your ISP yet, so it was normal that 0.0.0.0 would be shown on the status page.

One more thing, but incoming connections do work, so it has got lease from ISP.
Asus wont show it but it´s there

If i use wifi or wired connection inside my house, cannot reach any service outside, but incoming connection to ssh etc working normal.
But, factory reset it is

 

[Homie]

New firmware - new DHCP WAN issues

Just upgraded to firmware 3.0.0.4.374.5517 on my RT-AC66U.

It has previously run without issues for months, but right after upgrade I get intermittent loss of WAN IP, and get the same log message as others have seen:

"WAN Connection: ISP's DHCP did not function properly"

The state is persistent for almost an hour, and a restart does not appear to help. I have tried moving from Aggressive to Normal DHCP mode in my WAN settings but without any noticable difference...

Any ideas?

 

[rassi]

Now used 374.41 Merlin two days without an issue, hopefully issues wont come back. Used 30+30+30 reset before change to Merlin.

 

[L&LD]

Now used 374.41 Merlin two days without an issue, hopefully issues wont come back. Used 30+30+30 reset before change to Merlin.

Doing a reset before upgrading does not accomplish anything with regards to resetting the router to factory defaults. Neither does a 30+30+30 do anything with Asus routers as far as I know.

 

[BlowfishBoy]

I have Merlins 374.41 firmware installed and I also get the "WAN Connection: ISP's DHCP did not function properly" problem.
I'm on a fiber-optic ethernet network with a basic layer-2 switch installed as my operators CPE, so I do not have any cabel modem acting up in my set-up.
The connection to the ISP is a regular DHCP setup, so no PPPoE or anything else.

How I solve this problem is that I change to different cloned mac addresses, which for some reason resolves the problem for me. My ISP issues 15 minute leases, so this can be quite problematic for me when the problem appears.
Also, when I get the problem, the router shows the received IP-address as 0.0.0.0, but it shows displays my ISP's DNS servers correctly.

This was never an issue when I was on 374.39 and the only thing changed in my setup is the fimware release, so imho, this is a problem introduced in the latest firmware from Asus.

Merlin, is there anything I can provide you with so that you might get a grip of this problem??

Thanks in advance!

 

[rassi]

Doing a reset before upgrading does not accomplish anything with regards to resetting the router to factory defaults. Neither does a 30+30+30 do anything with Asus routers as far as I know.

Yes, it was peace of mind
I can be sure that i can´t blame ISP or Router setup for this.

Still going without an issue. Failoever not tested nor failback as main connection do work.

 

[usnyty]

First I need to thank RMerlin for providing such good firmware for us. I bought the AC68U because of your firmware mod.

After replace my old N16 to AC68U, I put the latest firmware 3.0.0.4.374.41 on it right out of the box. So all settings are from scratch.

After I done all my settings, it was running fine for a while, but I got disconnect the next day morning. The problem was exactly the same as BlowfishBoy. WAN got dropped, shows 0.0.0.0 , but still can connect from outside, just cannot go on internet from inside. The fastest way to get it back is to click the "Internet Connection" switch to flip it off, and turn it back on, then the WAN will get the IP back and everything will become normal.

This problem will happen at least twice a day. Here are some logs:

May 8 18:08:30 WAN(0) Connection: ISP's DHCP did not function properly.
May 8 18:08:30 stop_nat_rules: apply the redirect_rules!
May 8 18:08:36 dnsmasq[12230]: exiting on receipt of SIGTERM

May 8 19:01:43 rc_service: httpd 545:notify_rc restart_wan_if 0
May 8 19:01:43 dnsmasq[31141]: read /etc/hosts - 5 addresses
May 8 19:01:43 dnsmasq[31141]: read /etc/hosts.dnsmasq - 23 addresses
May 8 19:01:43 dnsmasq-dhcp[31141]: read /etc/ethers - 24 addresses
May 8 19:01:43 dnsmasq[31141]: using nameserver 71.250.0.12#53
May 8 19:01:43 dnsmasq[31141]: using nameserver 68.237.161.12#53
May 8 19:01:43 stop_wan(): perform DHCP release
May 8 19:01:43 kernel: Attempt to kill tasklet from interrupt
May 8 19:01:43 kernel: device eth0 left promiscuous mode
May 8 19:01:43 kernel: br0: port 1(vlan1) entering forwarding state
May 8 19:01:43 miniupnpd[31164]: ioctl(s, SIOCGIFADDR, ...): Cannot assign requested address
May 8 19:01:43 miniupnpd[31164]: Failed to get IP for interface eth0
May 8 19:01:43 miniupnpd[31164]: SendNATPMPPublicAddressChangeNotification: cannot get public IP address, stopping
May 8 19:01:43 kernel: device eth0 entered promiscuous mode

I never have this problem when I was using my Asus N16 with Victeck's Tomato firmware, so I pretty sure there is something wrong with the firmware.


Service Info:

Provider: Verizon FIOS New York
Plan: 50/25
Router: Asus AC68U

Please take a look when you have time. Thanks again for provide this firmware.

 

[L&LD]

Try turning off your router and modem/ONT for at least a good 10-15 minutes.

This will let the new router be properly registered with your ISP.

While all the settings are from scratch I would still recommend a full reset to defaults after the upgrade; which version did you update it from? I would still be doing a full reset to defaults to rule out any funky interaction with the old settings (defaults) that were available (or not).

Hope I'm on the right track for you.

 

[BluePhoenix]

Possible Solution?

Hey everyone,

I found this thread after I encountered the same issue with my router ASUS RT-AC66R. Did a firmware update yesterday and everything seemed to be fine and then it started disconnecting at random times... then I remembered I saw a similar issue before.

For some reason, after updating, this setting changed back even though I had modified it before. WAN > Internet Connection > DHCP query frequency and switch that from "Aggressive" to "Normal Mode".

From what I had researched months ago, it had to do with certain ISPs not allowing you to poll DHCP too frequently, so if the modem lost this setting and "aggressively" tried to get it back, the ISP would not fulfill the request.

My modem is a Motorola Surfboard, which I saw some of you mention earlier, so I hope this helps.

 

[tibere86]

Count me as another one affected by this. Happens every router reboot.

AC56U running Merlin's latest.
ISP = Comcast
DHCP Query set to NORMAL

 

[oc48]

same problem

I'm on firmware 374.41 (Asus RT-N66U) (former DD-WRT user for years)

everything ran fine for 2-3 months they yesterday internet went down. Did all the usual troubleshooting. never lost link with my Motorola SB6121 modem.

Could access the net from my windows laptop direct connected (Cat 6) to my modem. Never lost wifi from any of my mac computers, ipad, etc., to my router, but router WAN to Modem just would not come up. Talked to Comcast for hours....and nothing.

left cable modem unplugged overnight and same thing this morning. Just did a factory reset of my modem AND my router. running just fine now with no configs outside of WIFI done to my Router. I did get a message a week ago about too many updates from my DDNS provider.

I'm set on Normal DHCP settings for WAN.

Really suspect this is too many DHCP and/or DDNS requests from the router. Going to reconfigure everything again and will see how long it lasts this time.

I can't go through this again, and looking at the forums, it seems there is a bug, or a bad setting with to many requests for COMCAST........

I'd like to stay on this build but might have to go back to DD-WRT. I'll update again if any other changes occur.

 

[RMerlin]

I'm on firmware 374.41 (Asus RT-N66U) (former DD-WRT user for years)

everything ran fine for 2-3 months they yesterday internet went down. Did all the usual troubleshooting. never lost link with my Motorola SB6121 modem.

Could access the net from my windows laptop direct connected (Cat 6) to my modem. Never lost wifi from any of my mac computers, ipad, etc., to my router, but router WAN to Modem just would not come up. Talked to Comcast for hours....and nothing.

left cable modem unplugged overnight and same thing this morning. Just did a factory reset of my modem AND my router. running just fine now with no configs outside of WIFI done to my Router. I did get a message a week ago about too many updates from my DDNS provider.

I'm set on Normal DHCP settings for WAN.

Really suspect this is too many DHCP and/or DDNS requests from the router. Going to reconfigure everything again and will see how long it lasts this time.

I can't go through this again, and looking at the forums, it seems there is a bug, or a bad setting with to many requests for COMCAST........

I'd like to stay on this build but might have to go back to DD-WRT. I'll update again if any other changes occur.

The MAC blacklisting mostly occurs with Charter, not Comcast. I doubt that's your problem.

Remember to always do a 5-10 mins power off of your modem whenever moving it between a laptop and the router if you do any additional tests.

Try powering off the router for 30 secs and turning it back on to reset the electronics inside the router in case something caused the WAN port to get stuck. Also try swapping the WAN cable, just in case.

 

[pcbrand]

PPPOE serial link down - seems like router initiated not modem

I am experiencing the loss of WAN issue with the RT-AC68U and an ISP supplied VDSL modem setup in bridge mode. Have been monitoring this for the last couple of weeks and this seems to occur in time of high activity (ie. slamming the modem with huge amount of traffic).

Note that I have been using a different router before upgrading to the AC68U (A Linksys EA series) with exactly the same parameters and network topology and no issues were ever experienced. When the modem was hammered with traffic from the old router the connection responsiveness simply slowed down until the spike was over, but never disconnected like with the AC68U.

Hmm... since I know the network topology is bulletproof (been using this setup for the last 5 years with no issues) seems like the router is initiating the disconnect action - maybe because some modem keepalive flag is taking too long?

Not an expert on Linux or this router, but I noticed the following entry in the log, after which the router basically re-initiated the PPPOE session through the modem.

I am almost certain this is router initiated because the devices connected to the internet, which were streaming video at the time, did not experience an interruption until the actual moment the router decided that the link is disconnected and initiated a reset. In other words, the connection was up while the router thought it was down because the modem did not respond to 10 echo requests...

Jun 16 13:59:08 pppd[594]: No response to 10 echo-requests
Jun 16 13:59:08 pppd[594]: Serial link appears to be disconnected.
Jun 16 13:59:08 pppd[594]: Connect time 4307.6 minutes.
Jun 16 13:59:08 pppd[594]: Sent 994183538 bytes, received 1827900551 bytes.
Jun 16 13:59:08 miniupnpd[30756]: Failed to get IP for interface ppp0
Jun 16 13:59:08 miniupnpd[30756]: SendNATPMPPublicAddressChangeNotification: cannot get public IP address, stopping

Is there any way to increase this timeout (seems like 10 pings is the threshhold?) or disable this echo=request check altogether for testing purposes?


Thanks

 

[RMerlin]

I am experiencing the loss of WAN issue with the RT-AC68U and an ISP supplied VDSL modem setup in bridge mode. Have been monitoring this for the last couple of weeks and this seems to occur in time of high activity (ie. slamming the modem with huge amount of traffic).

Note that I have been using a different router before upgrading to the AC68U (A Linksys EA series) with exactly the same parameters and network topology and no issues were ever experienced. When the modem was hammered with traffic from the old router the connection responsiveness simply slowed down until the spike was over, but never disconnected like with the AC68U.

Try replacing the network cable between the router and the modem. If your previous router only had a 100 Mbits interface, it's possible that the cable you were using without any problem back then would not be stable with the Gigabit port the Asus has.

 

[wollyka]

Hi
I am also seeing this in my log and the internet is disconnecting.
But i am not sure if it is a problem with the modem loosing sync (DSL) or if the router is initiating the modem re-sync..

I am experiencing the loss of WAN issue with the RT-AC68U and an ISP supplied VDSL modem setup in bridge mode. Have been monitoring this for the last couple of weeks and this seems to occur in time of high activity (ie. slamming the modem with huge amount of traffic).

Note that I have been using a different router before upgrading to the AC68U (A Linksys EA series) with exactly the same parameters and network topology and no issues were ever experienced. When the modem was hammered with traffic from the old router the connection responsiveness simply slowed down until the spike was over, but never disconnected like with the AC68U.

Hmm... since I know the network topology is bulletproof (been using this setup for the last 5 years with no issues) seems like the router is initiating the disconnect action - maybe because some modem keepalive flag is taking too long?

Not an expert on Linux or this router, but I noticed the following entry in the log, after which the router basically re-initiated the PPPOE session through the modem.

I am almost certain this is router initiated because the devices connected to the internet, which were streaming video at the time, did not experience an interruption until the actual moment the router decided that the link is disconnected and initiated a reset. In other words, the connection was up while the router thought it was down because the modem did not respond to 10 echo requests...

Jun 16 13:59:08 pppd[594]: No response to 10 echo-requests
Jun 16 13:59:08 pppd[594]: Serial link appears to be disconnected.
Jun 16 13:59:08 pppd[594]: Connect time 4307.6 minutes.
Jun 16 13:59:08 pppd[594]: Sent 994183538 bytes, received 1827900551 bytes.
Jun 16 13:59:08 miniupnpd[30756]: Failed to get IP for interface ppp0
Jun 16 13:59:08 miniupnpd[30756]: SendNATPMPPublicAddressChangeNotification: cannot get public IP address, stopping

Is there any way to increase this timeout (seems like 10 pings is the threshhold?) or disable this echo=request check altogether for testing purposes?


Thanks

 

[pcbrand]

Try replacing the network cable between the router and the modem. If your previous router only had a 100 Mbits interface, it's possible that the cable you were using without any problem back then would not be stable with the Gigabit port the Asus has.

Thanks for quick reply RMerlin.

The previous router (Linksys EA2700) WAN port was also Gigabit interface...
Also, my ISP down speed is only 25Mb max so should never approach anything near gigabit speeds.

I'll give it a shot with a new cable anyways to eliminate it as a variable.

But just for testing purposes, I would have loved to be able to turn off this echo-request check....

 

[pcbrand]

Hi
I am also seeing this in my log and the internet is disconnecting.
But i am not sure if it is a problem with the modem loosing sync (DSL) or if the router is initiating the modem re-sync..

Hi wollyka

My ISP download speed is only 25Mb/s max - and this modem can handle speeds of at least 100Mb/s down so it's not like it is being overwhelmed by bandwidth. Also in my case I know the modem is not defective and doesn't lose sync since when I plug back my Linksys EA2700 router everything works great, and it has been working fine for years.

But to your point - the other thing that could possibly overwhelm the modem is when multiple connections are being created too quickly (AKA half-open connections) - I know the Asus router is probably faster in requesting/creating connections than pretty much anything on the market - is there any way to control the max number of new connections per second?

For example, Windows XP had a limitation of 10 new TCPIP connections per second because the older network cards could not handle faster requests.

 

[sybergeek]

Pointers :

1. Unplug Power & Cable from Modem > Unplug Power from Router > Wait 15 minutes > Plug in Cable and power to Modem > Wait 5 mins > Plug in Router.

2. Telnet to router and Execute following commands
nvram set detectWan=0
nvram commit
reboot

3. LAN > Switch Control > NAT Acceleration >> Disable it.

4. Check modem's Upstream Signal (192.168.100.1 > Signal @ Motorola SB6121) if it is 50 dBmV or higher then may be the splitter is bad quality or Junction box to home cable is poorly done. Call Internet cable guy to fix cable and bring signal level below 50 dBmV. Alternatively get RF Amplifier ( like Motorola Signal Booster BDA-S1 )

5. Put a cheap router between AC66U and Modem, to use AC66u as Access point.

6. Set Wireless > General > Channel Bandwidth > 40 Mhz

* I did #1, 2 & 3 so far it's been stable for past 7 hours. If I still face issue then I will try #6, then #5 and then finally #4. Mine is Comcast cable, Motorola SB6121 modem and AC66U router running on latest stock firmware.

Let try different combination of above and see what works the best.

Cheers !