AI Mesh Node Drops Offline - Constant DHCPREQUEST and DHCPACK messages

[abracadabra11]

Currently running setup noted in my signature (RT-AX3000 and RT-AC68U both on 384.19).

AI Mesh node throughput and connectivity has been great, but it appears to randomly drop Offline and then fail to reconnect. The log is littered with DHCPREQUEST and DHCPACK messages once the node falls offline. Restarting the main router fixes the issue temporarily until the node falls Offline again.

Jan  7 14:28:23 dnsmasq-dhcp[15414]: DHCPREQUEST(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX
Jan  7 14:28:23 dnsmasq-dhcp[15414]: DHCPACK(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX RT-AC68U-2818
Jan  7 14:28:46 dnsmasq-dhcp[15414]: DHCPREQUEST(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX
Jan  7 14:28:46 dnsmasq-dhcp[15414]: DHCPACK(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX RT-AC68U-2818
Jan  7 14:29:16 dnsmasq-dhcp[15414]: DHCPREQUEST(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX
Jan  7 14:29:16 dnsmasq-dhcp[15414]: DHCPACK(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX RT-AC68U-2818
Jan  7 14:29:46 dnsmasq-dhcp[15414]: DHCPREQUEST(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX
Jan  7 14:29:46 dnsmasq-dhcp[15414]: DHCPACK(br0) 192.168.68.99 XX:XX:XX:XX:XX:XX RT-AC68U-2818

Wireless log shows -63dBm, so signal should be plenty good.

Strangely, system log doesn't appear to have any entries before that noted below (even though the router has an uptime of 2+ days). The MAC noted in this entry is identical to the AI Mesh node, except the final digit is off by 1 (19 instead of 18). I don't have any devices with that MAC on my network. Is this maybe what's causing the constant DHCP messages? Is the WPA handshake occuring and being issued for the wrong MAC somehow?

Jan  7 14:28:19 hostapd: eth5: STA XX:XX:XX:XX:XX:19 RADIUS: starting accounting session
Jan  7 14:28:19 hostapd: eth5: STA XX:XX:XX:XX:XX:19 WPA: pairwise key handshake completed (RSN)

Any ideas where to start troubleshooting?

 

[L&LD]

First, I haven't seen an RT-AC68U that continued being stable in an AiMesh setup, eventually, it would drop (as you're seeing).

You may want to jump to 386.1 Beta 3 firmware for both routers, do a full reset, and see if AiMesh v2.0 is more stable for you then.

Some random thoughts below.

If you're using the same SSID's as you used in your old AC class network, it may be time to change that.

If you haven't removed the AiMesh node from the main router, performed a full reset on the RT-AC68U, and then re-added it, it may be worthwhile doing so now.

If you haven't fully reset the RT-AX3000 after flashing it to RMerlin firmware, that is where I would start.

Fully Reset Router and Network

New M&M 2020

How to Connect an AiMesh Node

L&LD | SmallNetBuilder Forums

 

[abracadabra11]

I'll definitely give consideration to a clean start on the RT-AC68U. But this was already done when I initially setup the network only a couple of weeks ago. Hoping to get some additional feedback on the WPA and DHCP behavior noted in OP.

 

[abracadabra11]

I manually assigned an IP address to the AI Mesh node and that appears to have resolved the issue for now. It's been nearly a week and the AI Mesh node has remained online since this change.

 

[abracadabra11]

I manually assigned an IP address to the AI Mesh node and that appears to have resolved the issue for now. It's been nearly a week and the AI Mesh node has remained online since this change.

Looks like I spoke too soon. Seeing same behavior again this morning.

 

[abracadabra11]

Anyone have any suggestions for troubleshooting? Since the DHCP messages flood the log, it doesn't seem like I'm able to look back far enough to see when the node actually drops offline.

 

[L&LD]

I would suggest testing with RMerlin firmware 386.1 Beta 5 today. Then following the suggestions in post 2.

I don't see the use of troubleshooting an older and very soon outgoing/forgotten firmware base.

At the very least, a full reset on both the main router and the AiMesh node using the same firmware (fully reset) is where I would begin.

 

[abracadabra11]

I would suggest testing with RMerlin firmware 386.1 Beta 5 today. Then following the suggestions in post 2.

I don't see the use of troubleshooting an older and very soon outgoing/forgotten firmware base.

At the very least, a full reset on both the main router and the AiMesh node using the same firmware (fully reset) is where I would begin.

I plan on testing once the version is final. But don't care to tinker with the beta at this stage. If this was a 384 beta then I would give it a shot on the node. But figure I'll create more problems by mixing 384.19 main router with 386 beta node given the notable changes to AIMesh code.

 

[Pergola Fabio]

@abracadabra11

Have you figured this out? My node goes offline too, when it happens i also see a flood of DHCP messages... I can still ping the node , all lan clients are still working...

How did you give the node a static IP or did you have it a DHCP reservation? I have mine nodes also set on DHCP reservation

 

[abracadabra11]

@abracadabra11

Have you figured this out? My node goes offline too, when it happens i also see a flood of DHCP messages... I can still ping the node , all lan clients are still working...

How did you give the node a static IP or did you have it a DHCP reservation? I have mine nodes also set on DHCP reservation

Used DHCP reservation and that seemed to fix the issue. The drop that I noted after the DHCP reservation may actually have been a power outage.

Ultimately, switching the AIMesh node to AP mode is what produced the best results in terms of throughout and stability. But this was only possible because I moved to a new home where MOCA became a possibility.

 

[Pergola Fabio]

i already have my nodes setup with DHCP reservation, think i'm going to drop aimesh...

how did you manually configure the AP? did you use channel 1-6-11 best practice? or all same channel? seems aimesh is setting all nodes on same channel

 

[abracadabra11]

i already have my nodes setup with DHCP reservation, think i'm going to drop aimesh...

how did you manually configure the AP? did you use channel 1-6-11 best practice? or all same channel? seems aimesh is setting all nodes on same channel

I'm using Auto for the channel. I previously had manual channels selected, but didn't see a noticeable difference when I moved to a new place where other WiFi networks aren't an issue.

 

[Pergola Fabio]

gonna try first asus official firmware on the node instead of merlin, if that doesnt help, i quit imesh

 

[Maloon]

gonna try first asus official firmware on the node instead of merlin, if that doesnt help, i quit imesh

Did you try Asus official firmware on the node?

 

[Pergola Fabio]

Yes , and now stable for 2 weeks now, i think it's fixed, let's hope

 

[Maloon]

Yes , and now stable for 2 weeks now, i think it's fixed, let's hope

Great! Can you tell me what firmware version are you using?

 

[Pergola Fabio]

Just latest stock

 

[L&LD]

Latest stock means little. Please specify the router and actual firmware used.

 

[Pergola Fabio]

Main router is ac88u running merlin , 386.3.2
Nodes are ac86u running stock , 386 build 44470