WAN (ISP DHCP) issues with 380.64_2 on RT-AC3200

[ManuCH]

I've been running 380.62_1 on my RT-AC3200 and it was very stable (uptime >100 days), zero issues.

Yesterday I thought I'd upgrade to 380.64_2. At first, routing didn't work at all (clients were not able to access the Internet, even though I was able to ping Internet IPs from the router's SSH console). I thought my Dual WAN setup was the culprit (I use load balancing mode), so I switched to failover mode. Then everything started working again.

A few hours later, I found the following message in the log:

Jan 25 08:52:16 WAN(0) Connection: ISP's DHCP did not function properly.

The WAN had failed over to my backup connection (LAN1), even though my WAN was working properly at that time. The only way to switch it back was to reboot the router.

By googling a bit I found someone recommending to change the WAN DHCP query frequency to "Normal mode", so I did, and now I'm waiting to see if it helps.

My questions:

• Is there a reason why load balancing dual WAN is not working anymore? I know Merlin doesn't touch the dual WAN code, but is it known if Asus changed something in the GPLs used between 380.62_1 and 380.64_2?
• Why would I get this "ISP's DHCP did not function properly" message? Was anything changed in the DHCP client logic between those 2 versions?

Unrelated: I get the "Your router is running low on free NVRAM" message. Usage is currently 62727 / 65536 bytes. Is that "normal"? Do I need to clean up something? How?

 

[Piter]

I've been running 380.62_1 on my RT-AC3200 and it was very stable (uptime >100 days), zero issues.

Yesterday I thought I'd upgrade to 380.64_2. At first, routing didn't work at all (clients were not able to access the Internet, even though I was able to ping Internet IPs from the router's SSH console). I thought my Dual WAN setup was the culprit (I use load balancing mode), so I switched to failover mode. Then everything started working again.

A few hours later, I found the following message in the log:

Jan 25 08:52:16 WAN(0) Connection: ISP's DHCP did not function properly.

The WAN had failed over to my backup connection (LAN1), even though my WAN was working properly at that time. The only way to switch it back was to reboot the router.

By googling a bit I found someone recommending to change the WAN DHCP query frequency to "Normal mode", so I did, and now I'm waiting to see if it helps.

My questions:

• Is there a reason why load balancing dual WAN is not working anymore? I know Merlin doesn't touch the dual WAN code, but is it known if Asus changed something in the GPLs used between 380.62_1 and 380.64_2?
• Why would I get this "ISP's DHCP did not function properly" message? Was anything changed in the DHCP client logic between those 2 versions?

Unrelated: I get the "Your router is running low on free NVRAM" message. Usage is currently 62727 / 65536 bytes. Is that "normal"? Do I need to clean up something? How?

Yesterday i connect my modem Huawei E3276s Hilink and setup in router DUAL WAN with FAIL OVER. WAN1 is my ISP with modem cabel DHCP and WAN2 is modem USB. I also have this ISP's DHCP did not function properly when i test working DUAL WAN. Sometimes on WAN1 sometimes on WAN2

I use FW 380.65 beta2

 

[ManuCH]

Yesterday i connect my modem Huawei E3276s Hilink and setup in router DUAL WAN with FAIL OVER. WAN1 is my ISP with modem cabel DHCP and WAN2 is modem USB. I also have this ISP's DHCP did not function properly when i test working DUAL WAN. Sometimes on WAN1 sometimes on WAN2

I use FW 380.65 beta2

Have you already tried a factory reset and restore settings using the script? I'm pondering on whether that might be necessary for some reason.

 

[Piter]

Have you already tried a factory reset and restore settings using the script? I'm pondering on whether that might be necessary for some reason.

No. I dont done factory reset. But what script you have mean ?

 

[ManuCH]

No. I dont done factory reset. But what script you have mean ?

I mean the script I linked: https://www.snbforums.com/threads/user-nvram-save-restore-utility-r24.19521/

This simply avoids you having to re-do the config from scratch, but maybe it will also copy over whatever is causing the issue. I'm not sure. Maybe someone else will chime in.

 

[KtBken]

You don't say who your ISPs are but i have had a similar issue with virgin media in the uk. This seems to be an issue at there end as i have never had it with any other provider. Setting dhcp to normal mode has reduced the frequency of the issue and it will now get an address although it sometimes takes a few minutes.
I have a 5300 running in duel wan mode with load balancing on 65.beta2 which seems to be running ok.

 

[ManuCH]

You don't say who your ISPs are but i have had a similar issue with virgin media in the uk. This seems to be an issue at there end as i have never had it with any other provider. Setting dhcp to normal mode has reduced the frequency of the issue and it will now get an address although it sometimes takes a few minutes.
I have a 5300 running in duel wan mode with load balancing on 65.beta2 which seems to be running ok.

My ISP is UPC Switzerland and the cable modem is set to bridge mode (so I'm getting the IP directly from their central office).

I find it interesting that the issue only started appearing with 380.64_2, that's why I'm asking if anyone (@RMerlin ?) knows whether something has changed in the DHCP client and/or Dual WAN code between 380.62_1 and 380.64_2.

 

[niraviry]

I am having the same issue but with RT-AC88U. My Primary WAN is VDSL from external modem and the failover WAN - Huawei K5150 (unlocked Vodafone) used with my cellular ISP.
When I have tested it and disconnected the primary WAN, the K5150 took over but when I returned it back it did not. I had to disable dual WAN to make it return to the primary WAN (caused a reboot).
It was the 1st time I use Merlin F/W and the router was reset so it is fresh.
With the E398 which is not HiLink modem it works OK.

 

[Finch]

Same here, AC68U randomly reports dhcp issues with my isp, connected to Cisco cable modem.

 

[ManuCH]

Same here, AC68U randomly reports dhcp issues with my isp, connected to Cisco cable modem.

Has this only started with 380.64_2, or did you have the issue previously as well?

 

[Finch]

It was fine before, started with 380.64_2.

 

[niraviry]

I use it the 1st time. There is a 65_Beta 2 available now. I will wait for the release.

 

[Piter]

I still have problem with dhcp on 380.65 beta 2. I use fatal dual wan with usb 4g and problem dhcp sometime is on wan sometimes on usb. I dont know what is a problem. But asus still have a problem with asus wrt. Ehhhh and nothing repaire.

 

[Finch]

Went back to 380.64_1 and it's fine. BTW I do not use dual WAN.

 

[niraviry]

Piter.

I totally agree with you. I am very disappointed with Asus support. I had an issue for few weeks with dual WAN. Asus routers are sold in my country so I have bought it in NY. One thing I can say in favor of Asus - they have a support in my country to the routers although thay do not sell them. The support is slower that a turtle. There were 2 issues. One that is part of this topic and the 1st basic one was a non working dual WAN. Eventually I solved it (see the topic about HLink modems). It is working but totally unstable. The Asus respond to the non working dual WAN was a hardware problem. Actually it was documentation problem.

 

[Dezzo]

I too have had the "Your ISP's DHCP server isn't functioning properly" error message.

In my scenario, I recently had my ISP modem upgraded and have an AC68 router connected behind it.
DMZ doesn't work and router pulls a private LAN IP instead of a public WAN IP as it should, creating double NAT.

While I still believe it's the poorly written firmware on my ISP's modem having broken DMZ mode, perhaps it's an AC68 firmware bug?