Asus RT-AC88U merlin 386.1_2 getting drops on Verizon FIOS "WAN_Connection: ISP's DHCP did not function properly."

[deathbyburk]

Latest stock firmware with "WAN issues fix" and almost 24 hours no problems. I could not make more then 3 at any time. I'll keep posted but so far so good

 

[Morris]

Latest stock firmware with "WAN issues fix" and almost 24 hours no problems. I could not make more then 3 at any time. I'll keep posted but so far so good

That's interesting. To implicate the Merlin code, you will need to allow a few more renewals and/or force them. Then you need to go back to the Merlin code and see if the problem persists.

Morris

 

[deathbyburk]

So 2 full days and no issues. Not a merlin issue but was an ASUS until they patched 2 patches ago(assumption). Merlin should be merging that code next/soon.

 

[Morris]

So 2 full days and no issues. Not a merlin issue but was an ASUS until they patched 2 patches ago(assumption). Merlin should be merging that code next/soon.

This assumption should be proven. Revert to the Merlin firmware and/or equivalent Asus version and see if the problem comes back. So far you have proven that a reset to factory resolves your issue.

Morris

 

[deathbyburk]

oh sorry maybe I was not clear. gigabit/fiber is very new and not sure what/if is different then my prior 2 cable internet providers I have been using since I upgraded from my u.s. robotics 56k but could not stay connected. very sinilar "connected to internet but no internet" issues I see posted.


386.2_2
issues
clean reset - issues

386.2_4
same

384_19
clean reset and lost net after few hours and honestly had less patience at this point to mess with settings as I did other

Latest stock (not searching version atm)
3 days ROCK solid

 

[Morris]

oh sorry maybe I was not clear. gigabit/fiber is very new and not sure what/if is different then my prior 2 cable internet providers I have been using since I upgraded from my u.s. robotics 56k but could not stay connected. very sinilar "connected to internet but no internet" issues I see posted.


386.2_2
issues
clean reset - issues

386.2_4
same

384_19
clean reset and lost net after few hours and honestly had less patience at this point to mess with settings as I did other

Latest stock (not searching version atm)
3 days ROCK solid

Now that you are good for 3 days, load Merlin 386.2_4 with a full reset. Then configure manually and see how it works.

Good luck,

Morris

 

[OxqIMJ]

Is there any further updates on this? I am using the latest stock firmware and have this exact problem.

 

[bilged]

I'm on 386.2_6 and no longer have the problem. Looks like it's been fixed. I did do a clean update though too.

 

[Morris]

Is there any further updates on this? I am using the latest stock firmware and have this exact problem.

The latest firmware allows you to set "DHCP query frequency" to Continuous. The setting can be found from the left side tab WAN and top tab Internet Connection. It is the third from the bottom. I'm on FIOS and rock solid.

Morris

 

[sbsnb]

I've been using FiOS for the past 11 years with whatever is the default DHCP setting and never had an issue.

 

[Morris]

I've been using FiOS for the past 11 years with whatever is the default DHCP setting and never had an issue.

Is your Asus router connected directly to the ONT? Many people have had issues and it lead to Asus adding the continuous setting which appears to have corrected the issue. My connection only changed to the ONT recently and I chose continuous when I connected to avoid the reported issues so I'm not a use case for this fix yet have no issues with continuous selected. Possibly the model of the ONT you have makes a difference.

Morris

 

[sbsnb]

Is your Asus router connected directly to the ONT?

Yes, by Ethernet cable. I don't have TV or phone service through FiOS.

 

[OxqIMJ]

The latest firmware allows you to set "DHCP query frequency" to Continuous. The setting can be found from the left side tab WAN and top tab Internet Connection. It is the third from the bottom. I'm on FIOS and rock solid.

Morris

thanks will give it a try.

 

[OxqIMJ]

Just installed the latest firmware released today (Version 3.0.0.4.386.45934) and changed the setting to continuous and it still disconnects

 

[Morris]

Just installed the latest firmware released today (Version 3.0.0.4.386.45934) and changed the setting to continuous and it still disconnects

I'm running the latest Merlin code with a RT-AX86U on FIOS connected via cat 5 to the ONT and no troubles. Try changing cables. The other possibility is something is defective: ONT or router.

Good luck,

Morris

 

[k0rp]

I posted this in another thread, but this one appears to be almost identical to my issues. Here's what I've tried over the past few weeks:

Decided to bite the bullet after trying just about everything else (DHCP query intervals, various admin settings, replacing ethernet cords, etc.) and performed a downgrade to Merlin 384.19. I will report back if I keep experiencing any DHCP issues from with my RT-AC68U but so far so good.

My current network setup for anyone else trying to troubleshoot this issue:
Verizon FiOS internet (ONT) ---- Cat 6 ----> RT-AC68U (running Merlin 384.19) ----> switch(es) ----> pihole

EDIT: 1 full day since firmware downgrade. No issues yet and I already have noticed my IP change at least 1 time.

EDIT #2: I'm now noticing my upload speeds drop from 800-ish Mbps down to under 100 Mbps with no change to download speed. No issues with actual internet access yet.

EDIT #3: I restarted both my PC and laptop (which in the diagram above, sit at the same topology as the pihole) ONLY and my upload speeds are back to normal. I did not restart/reboot my router or ONT, it just magically fixed itself?

EDIT #4: Just lost connectivity (roughly 2 days of uptime). Router indicated that I was still connected to the internet, with 1hr 50 minutes left on my DHCP lease. Tried manually restarting WAN via router, which resulted in the "Your ISP's DHCP does not function completely". Rebooted router, connectivity restored, same IP address as before I lost connectivity, but now I'm seeing much slower upload speeds again on several devices even after restarting them. Performed an additional router reboot and upload speeds seem back to normal on all devices. New IP address acquired during second reboot.

EDIT #5: Got on the horn with a Verizon agent, asking them to take a look. They did a "refresh" of my internet, which required another router reboot and it seems like I got another fresh IP. After that they said that since I'm connected to the internet and everything on their end looks good, that they can't do anything else. This is so much fun.

EDIT #6: After that talk with Verizon yesterday, I turned off my router for a solid 30-45 minutes and went for a walk. When I returned, I turned my router back on and it's been pretty solid since. Let's see how long this lasts (16ish hours so far).

EDIT #7: No full internet connectivity drops, yet (36+ hours). I am noticing that sometimes my internet speeds (from what I can tell, mostly upload) slows down to a snail's pace, sometimes as low as 1Mbps. I have a work laptop that's connected to my network (wired) and it seems to lose VPN connectivity when this happens, but it is able to reconnect shortly. During this time, I can still access the internet, but it feels like something is going on. My IP seems to be the same the entire time.

EDIT #8: After experiencing the above with what appears to be speed degradation, after 10 minutes we're back up to 100% speeds. Bizarre.

EDIT #9: (6 days): I thought it was finally "fixed". Nope. Internet just cut out and a router reboot didn't fix it, as it started displaying the ISP DHCP issue on the home screen. Rebooted the router again, finally have internet again.

EDIT: #10 (15 hours): Another disconnect. Reboot of router was the only way to get back online, same IP address. I did see some upnp error logs so I disabled that for now to see if that helps. I'm currently using the Continuous setting for the DHCP query frequency.

 

[drinkingbird]

I posted this in another thread, but this one appears to be almost identical to my issues. Here's what I've tried over the past few weeks:


EDIT: #10 (15 hours): Another disconnect. Reboot of router was the only way to get back online, same IP address. I did see some upnp error logs so I disabled that for now to see if that helps. I'm currently using the Continuous setting for the DHCP query frequency.

You're still having disconnects after going down to 384.19? Did you do a full nuclear downgrade and not restore your settings (re-input them manually)?

I had this issue on 386 (don't recall which one I tried) and did a clean/nuclear downgrade to 384.19 and it has been fine ever since (other than the GUI becoming unreachable from time to time which was a common bug with that version). I've been waiting until several reports that 386 has fixed this until trying again. I tried continuous mode etc all without avail. Connected directly to ONT using a short CAT5E cable. This is on RT-AC68U.

Someone above mentions 386.2_6 working but not sure if they're just lucky or if it has really been fixed. It seems FIOS has different lease times in different areas. I get 2 hours leases and those seem to be the ones with the most issues. From what I can tell, the router is not properly sending DHCPACK to FIOS server so the IP gets given to someone else and thus you are blocked until you force a new IP. Was just a guess, I wasn't motivated enough to put a sniffer on the WAN to see.

If you've done a clean downgrade and are still having issues (especially if you are seeing speed dropouts) it may be a different issue. Connectivity issues on the WAN will also result in DHCP errors.

Not sure if anyone with 386.3 that was having the issue is now ok. Probably wait for 4 or 5 to come out before trying, honestly 384 is working fine for me other than having to reboot sometimes due to GUI freezing up.

 

[Morris]

I'm on FIOS, my RT-AX86U is connected to the ONT via CAT 5E and my DHCP lease is 2 hours. I'm Merlin V 385.3.2 and my DHCP query frequency is Continuous Mode. I do not have issues.

I feel you need to investigate why the DHCP handshake is falling during normal operation. I recommend trying:
- Replace jumper cable between ONT and router
- Run a speed test and see if you are getting at least the speed you are contracted for in both directions
- Consider your network load. If your link is heavily loaded with fire transfers, torrent's, or high packet rates from things such as gaming, you need to moderate the load.
- If it's file transfers, try CAKE QOS
- If it's packet rate, your router is under powered and you need an upgrade

Good luck,

Morris

 

[k0rp]

You're still having disconnects after going down to 384.19? Did you do a full nuclear downgrade and not restore your settings (re-input them manually)?

I had this issue on 386 (don't recall which one I tried) and did a clean/nuclear downgrade to 384.19 and it has been fine ever since (other than the GUI becoming unreachable from time to time which was a common bug with that version). I've been waiting until several reports that 386 has fixed this until trying again. I tried continuous mode etc all without avail. Connected directly to ONT using a short CAT5E cable. This is on RT-AC68U.

Someone above mentions 386.2_6 working but not sure if they're just lucky or if it has really been fixed. It seems FIOS has different lease times in different areas. I get 2 hours leases and those seem to be the ones with the most issues. From what I can tell, the router is not properly sending DHCPACK to FIOS server so the IP gets given to someone else and thus you are blocked until you force a new IP. Was just a guess, I wasn't motivated enough to put a sniffer on the WAN to see.

If you've done a clean downgrade and are still having issues (especially if you are seeing speed dropouts) it may be a different issue. Connectivity issues on the WAN will also result in DHCP errors.

Not sure if anyone with 386.3 that was having the issue is now ok. Probably wait for 4 or 5 to come out before trying, honestly 384 is working fine for me other than having to reboot sometimes due to GUI freezing up.

Yep, full "nuclear" downgrade and re-entered all settings manually. My lease appears to be 2 hours as well.

I'm on FIOS, my RT-AX86U is connected to the ONT via CAT 5E and my DHCP lease is 2 hours. I'm Merlin V 385.3.2 and my DHCP query frequency is Continuous Mode. I do not have issues.

I feel you need to investigate why the DHCP handshake is falling during normal operation. I recommend trying:
- Replace jumper cable between ONT and router
- Run a speed test and see if you are getting at least the speed you are contracted for in both directions
- Consider your network load. If your link is heavily loaded with fire transfers, torrent's, or high packet rates from things such as gaming, you need to moderate the load.
- If it's file transfers, try CAKE QOS
- If it's packet rate, your router is under powered and you need an upgrade

Good luck,

Morris

I will try replacing as much of the cable as possible to see if that helps. With my current setup, I almost always get advertised download speeds with slightly lower upload speeds, with some periods of upload speed degradation around the time of a DHCP lease renewal. It eventually comes back up though. My WAN usage is pretty "normal" -- video streaming services, internet browsing, and gaming. I have only experienced one full disconnect during gaming and that was probably a month ago before I downgraded to 384.19.

EDIT #1: Replaced some ethernet (CAT 5E) between ONT and Router and noticed that I have already picked up a new IP address, after appearing to keep the same IP address for 11 days.

EDIT #2: Seeing a new log message repeatedly happening for the same MAC address.: Nov 21 12:03:44 syslog: WLCEVENTD wlceventd_proc_event(500): eth1: Auth [MAC ADDRESS], status: Successful (0) . Tried a router reboot and they're still flooding in. I upped the log level to avoid this, but any ideas what's causing it? The MAC address doesn't appear to be any of my devices so I'm not sure what's going on.

EDIT #3: Turned off WPS and the above log message has stopped and I no longer see the MAC address in question in the Wireless Log tab of System Log tab.

EDIT #4: My work VPN dropped for a moment and I'm experiencing slow upload speeds (as noted before). Replacement of the ethernet between ONT and Router did not fix my issue.

EDIT #5: I kinda gave up -- wanted to try out MoCA since the house I am in has coax run everywhere and it would be painful for trying to run cat6 with how everything is laid out. Got a TP-Link wired router to place at my electrical box to route everything and turned my RT-AC68U into a AP with a ethernet backbone. Purchased a pair of MoCA adapters on sale for black friday and have it all wired up and running. Works nice if anyone was thinking about trying it out.

 

[OxqIMJ]

After giving up on this, I saw some new firmware updates for my RT-AC68U and pulled it out of the closet. It appears this problem is fixed. On the latest Asus firmware and have not had a single drop connection in a few weeks.