Frequent disconnects with third-party router (GL.iNet Flint 2) on FTTP

[Prawncocktail]

I use a third-party router for my EE FTTP connection. The router is a GL.iNet Flint 2. The connection drops frequently.

My current settings are:

• Protocol: PPPoE

• Username: [email protected]

• Password: [empty]

• MTU: 1492

Please give me the correct configuration data to make a stable connection. Please supply the correct values for:

• Username

• Password

• MTU

• VLAN ID

• Authentication protocol

Thank you.

[JimM11 · Accepted solution]

@Prawncocktail You have them all as per the third party guide. Just select your correct connection type for you it's Full Fibre.

How Do I Use My Own Router for EE Home Broadband? | Broadband Help | EE

[XRaySpeX]

You're already correct! Don't set a VLAN ID. It's only for FTTC.

Is it the BB as a whole dropping or just the WiFi connections? Do Ethernet connected devices stay working? What colours are the light sequence when it happens?

BB issues are rarely widespread but specific to your own line.

If it is a neighbourhood issue, you could consult Check service in your area  to see if any faults are being attended to in your local area. You may also Report a Problem there.

Have you reported your problem with your BB to EE? They can't do anything about it until you do. Text HELP to 66033 to get EE to test the line & see what needs doing.

[Prawncocktail]

WiFi is solid, just seems to be an issue with the connection dropping then establishing connection quite quickly soon after. I set up a docker container to ping 1.1.1.1 to keep an eye on connectivity after making some changes. Here is a bit more detail on what I have set up. 

• Router Hardware: Upgraded from the standard EE Hub to a GL.iNet Flint 2 router connected directly to the Openreach ONT via Ethernet.

• WAN Protocol: Configured as PPPoE on the WAN interface.

• Authentication: Using the standard broadband username [email protected] with BT as the PAP/CHAP password.

• MTU Size: Overridden and adjusted down from the default 1500 to 1492 to account for PPPoE packet overhead.

• Link Control Protocol (LCP): Adjusted LCP echo failure threshold to 5 and interval to 30 seconds to handle keep-alives and prevent premature drops.

• Hardware Acceleration: Enabled hardware acceleration (flow offloading) to support multi-gigabit throughput.

[JimM11]

@Prawncocktail Lower your LCP interval down from 30 seconds to 5 and the failure up to 10, 

What indication are you getting that the WAN is dropping from the Flint hub? You may find some useful info on your Hubs Forum if you are fighting Wan drops but you do not say so making it hard for anyone else to work out.

https://forum.gl-inet.com/

[Prawncocktail]

I'll give that a shot, thanks. 

The internet drops out, I can't connect to anything outside my network. I'm not sure how else to describe it tbh.

Here you can see clearly how the connection drops out at 2026-09-07 19:15:46, it then establishes the connection a minute later. 

[JimM11]

@Prawncocktail Will take a look at your posted pictures if they get cleared for general viewing, are you new to EE as a service and started straight out with your Flint2 Router.

What is your history with it, ie worked great started doing this, did a FW update it all went to pot, switched on iPv6 and now it does not play, anything helps.... ie stuck the EE Hub on it works as expected?

Do you know how to check your wan IP that EE issues to you either/both iPv4/6 depending on what you have setup on the flint? Think the EE was very sticky so did not change much even if the wan got dropped for a little time.

[XRaySpeX]

So, 1 discon in 3 hours doesn't make a summer! How frequently is "Frequent" that you are experiencing?

Don't worry, the images were awaiting approval. I have now done so & they should appear soon.

[DoingItSideways]

Look at the router's system log to find out what's happening during these drops.

[JimM11]

@Prawncocktail You pictures are showing at the time your Flint 2 had an issue OR there was possibly an Internal/external problem and as @DoingItSideways points out have a look at your hub's log file if it's a kernel panic on the Hub the log will fill up with many entries and they will be Time stamped as off the time, all off the other information you have posted setup wise would appear to be correct as you are online connected but having a BLIP, also what is the boot time for your hub that's a good starting place, my Asus takes a lot longer than a minute and half to boot, but a kernel panic is way quicker might have a capture somewhere will have a look later does not happen very often normal 24hour logs for day's are extremely short.

Typical for a log on an asus it was back months ago though.... ""This might have been a VF powerhub switch in from the sky unit as did not have time to mess with the ISP changeover due to 3 week business trip leaving noon that day.""

Apr 15 06:20:15 kernel: eth0 (Int switch port: 5) (Logical Port: 5) (phyId: 19) Link DOWN.
Apr 15 06:20:20 WAN Connection: WAN(0) link down.
Apr 15 06:21:08 kernel: eth0 (Int switch port: 5) (Logical Port: 5) (phyId: 19) Link Up at 1000 mbps full duplex
Apr 15 06:21:10 WAN Connection: WAN(0) link up.

In between all the entries and have skipped until connection was re-established and up..... You will see that was a 10minute outage so suspect OR drop exchange etc.... At 6.20am would not be me for sure that early in the morning. There was 6 down/up events happened during the period until it was established and all working.

Apr 15 06:32:02 WAN Connection: WAN was restored.
Apr 15 06:32:03 wan_up: Find a 2.5G port on [eth0]
Apr 15 06:32:04 ping_target_with_size: Ping test is complete.
Apr 15 06:32:06 ping_target_with_size: Successful to ping target(192.168.1.1) with size(79)
Apr 15 06:32:06 wan_up: Restart DDNS