Cannot use BQM with Hyperoptic

I am on Hyperoptic with a static IP (IPv4) address. BQM gives always 100% packet loss (https://www.thinkbroadband.com/broadband/monitoring/quality/share/create/705f05a47946ad83fb0ce8697f91f4cee173504c)

I know that the Hyperoptic router does not respond to ping requests, but I get the same results running Openwrt on a Raspberry Pi 4 either straight into the fibre socket or connected to the Hyperoptic router. In the latter case the Raspberry is on DMZ. Either way the Raspberry responds to ping request from outside the network (e.g. from my 5G phone network) but the BQM returns a solid red reading.

Is Hyperoptic doing something strange? Is there anything I can do about it? I am experiencing very frequent dropouts (hence my need to document the connection quality) but not 100% packet loss!

TIA

I have a TBB BQM pointed to a Hyperoptic FTTP connection (also with static IP). It's been stable. This is running to my own router.

What IP range have you been assigned to? I am on 140.228.71.0/24

My Hyperoptic BQM (live)

Edit: the hyperlink in your OP is not working. Can you re-share the direct link made from "share live graph" on the main BQM page.

Edited by Pheasant (Tue 06-May-25 17:51:02)

Thanks for your reply. My IP address is 140.228.71.156.

My Hyperoptic BQM live

At the moment the Raspberry/Openwrt is connected to the Hyperoptic router in the DMZ. If I connect it directly to the socket I get the same results but the connection drops after 20 minutes or so...

Edited by Pippo123456 (Tue 06-May-25 18:56:08)

Your IP address is definitely reachable. Try deleting the BQM completely and re-create it from scratch.

Also setup a ping based monitor at f8lure.mouselike.org and see what overnight results you get from there.

I have created new BQM

and setup a F8lure account
F8lure ping results

On both 100% packet loss. It is as if I am behind a CGNAT despite a static IP!

Thanks for all your help!

Edited by Pippo123456 (Tue 06-May-25 19:47:41)

Is this a Hyperoptic business or residential service?

Humour me - can you confirm what’s returned by visiting say https://whatismyipaddress.com/ ?

Edited by Pheasant (Tue 06-May-25 20:05:12)

Residential.
Both Ip.me and whatsmyipaddress.com return my static IP.

It is very weird.

Have you got the Hyperoptic router still in play currently?

Yes

Remove the Hyper router and connect another router directly to the ONT.

I will try again tomorrow (the span is currently watching the Champions League). I tried it multiple times, the last one this afternoon, and BQM always reports 100% packet loss.

Sorry, it took me a while as everybody was busy online.

I connected the openwrt/raspberry router directly to the ONT just before 15:00 today. Before that the openwrt was connected to the Hyperoptic router.

Here you can see the results from f8lure and AAISP

Still full red on a new BQM

The AAISP graph is the only one that shows some sign of life after 15:00 today but its text counterpart reports 100% packet loss.

At a certain point (let's say 16:00) I will move back to using a VPN which is the only way I have a stable connection. Results after that time will be uninformative.

Not sure if it will be useful, but have you tried tbb's traceroute tool? If it can't see your router, you might at least find out where...

How are you assigning the WAN side address to the router - are you using the DHCP client or manually assigning the IP, gateway address?

DHCP client. I can access the internet and the router responds to ping requests to 140.228.71.156

1 thinkbroadband-gw9.core-rs2.thw.ncuk.net 15 15 1ms 1ms 1ms
2 te1-52-37.core-rs1.thw.ncuk.net 15 15 1ms 1ms 3ms
3 xe-0-1-6-38.edge-rt1.thn.ncuk.net 15 15 0ms 1ms 6ms
8 71.228.140.156.bcube.co.uk 15 15 1ms 1ms 1ms

This is what I get. Sorry for the poor formatting.

I'm no expert so doesn't tell me much, except to wonder what happened to steps 4 to 7?

Anyone else?

Is the assigned gateway 140.228.71.1 ?

Those intermediate nodes are not responding to the packets from the traceroute in a timely manner (or at all). Normal.

The Luci interface does not show the gateway address but this is what I get from the command line

ip route show
default via 140.228.71.1 dev eth0 src 140.228.71.156
140.228.71.0/24 dev eth0 scope link src 140.228.71.156
192.168.1.0/24 dev br-lan scope link src 192.168.1.1

From my limited understanding, I'd say yes.

Edited by Pippo123456 (Thu 08-May-25 17:04:06)

OK, thanks- haven't seen that before, must just have been lucky.

Yep same gateway as my HO connection, unsurprisingly given we're on the same subnet.

Just to clarify, on testing this evening from my connection:- I can only successfully ping your address from my own Hyperoptic WAN connection. However if I try from my other WAN connection then it fails. I've also tried from another router running an EE connection and that also fails to your IP address.

So I believe the issue is within the Hyperoptic setup of your static IP address.

That is likely because, as I wrote previously, from 4pm today I am again connected to a VPN from the router which is the only way to obtain a stable connection.

I cannot ping my HO IP from my phone 5G connection whenever I am connected to the VPN, but I can when the router is not using the VPN. So can you according to your post of Tue 06-May-25 19:16:28.

Thanks!

Edited by Pippo123456 (Thu 08-May-25 18:46:51)

Now you should be able to ping my IP from anywhere. In case you are still online

Edit. Scratch that last post 🤣

Edited by Pheasant (Thu 08-May-25 18:56:07)

Do you have a business or personal contract? I have no possible explanation for the different behaviour of our two connections wrt BQM.

Also do you know of any Linux utility that provides the same kind of analysis of BQM but pinging from my router?

I have set up a cron job to run the Speedtest cli at 5 minute intervals and was able to document the dropouts but something that produces a graphical output without further manipulation would be an improvement.

Business HO connection.

Try creating a reverse test at f8lure. Instructions are here:
https://f8lure.mouselike.org/index.asp?DoAction=Reve...

In reply to a post by Pippo123456:

… I am again connected to a VPN from the router which is the only way to obtain a stable connection.

I know you’ve been trying to reach out to HO, but worth persisting to get through to second or third line support on this matter. There’s definitely something awry with your connection and its setup.

I suspected that!

They are clearly doing something non-kosher with the non-business connections.

I have tried the f8lure reverse test from a wired connection to the openwrt router connected directly to the OTN.

If I connect through the VPN it works fine. If I connect without the VPN:

1. The graph on f8lure never shows anything.
2. This is what I get in the command window from which I have launched the test

1746779541 19 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779545 29 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779547 37 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779551 47 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779553 55 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779557 65 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779559 73 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779563 83 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779565 91 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779569 101 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779571 109 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779575 119 windows.mouselike.org Lost Ping (current loss 1.02%)
Posting to central logger.
1746779577 11 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779583 23 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779589 35 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779595 47 windows.mouselike.org Lost Ping (current loss 1.02%)
ERROR AT windows.mouselike.org
System.Net.WebException: The operation has timed out.
at System.Net.HttpWebRequest.GetResponse()
at System.Net.WebClient.DownloadBits(WebRequest request, Stream writeStream)
at System.Net.WebClient.UploadBits(WebRequest request, FileStream readStream, Byte[] buffer, Int32 chunkSize, Byte[] header, Byte[] footer)
at System.Net.WebClient.UploadValues(Uri address, String method, NameValueCollection data)
at reversef8lure.Program.UpdateFilesAppSub() in \\VBOXSVR\reversef8lure\reversef8lure\Program.vb:line 121
at reversef8lure.Program.NewPingTest() in \\VBOXSVR\reversef8lure\reversef8lure\Program.vb:line 85
ERROR AT windows.mouselike.org
1746779601 59 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779602 66 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779607 77 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779608 84 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779613 95 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779614 102 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779619 113 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779620 120 windows.mouselike.org Lost Ping (current loss 1.02%)
Posting to central logger.
1746779625 11 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779631 23 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779637 35 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779643 47 windows.mouselike.org Lost Ping (current loss 1.02%)
1746779649 59 windows.mouselike.org Lost Ping (current loss 1.02%)

forever!

I bookmarked your existing f8lure BQM and on the secondary test (hosted from AAISP) I can see there are two windows where that BQM appears to be ‘normal’
- yesterday between approx 14:30 and 17:00
- this morning between 7:30 and 9:30

Presume this is when you had your tunnel running or something else?

yesterday was through the VPN, today between 7.30 and 9.30 without. I briefly tried again with the VPN after writing my previous reply and again. Below is the terminal output of the test in the latter case: all good

Configured with test secret: REMOVED
1746780420 0 windows.mouselike.org response in 2ms
1746780421 1 windows.mouselike.org response in 2ms
1746780422 2 windows.mouselike.org response in 2ms
1746780423 3 windows.mouselike.org response in 2ms
1746780424 4 windows.mouselike.org response in 2ms
1746780425 5 windows.mouselike.org response in 2ms
1746780426 6 windows.mouselike.org response in 2ms
1746780427 7 windows.mouselike.org response in 2ms
1746780428 8 windows.mouselike.org response in 2ms
1746780429 9 windows.mouselike.org response in 2ms
1746780430 10 windows.mouselike.org response in 2ms
1746780431 11 windows.mouselike.org response in 2ms
1746780432 12 windows.mouselike.org response in 2ms
1746780433 13 windows.mouselike.org response in 2ms
1746780434 14 windows.mouselike.org response in 2ms
1746780435 15 windows.mouselike.org response in 2ms
1746780436 16 windows.mouselike.org response in 2ms
1746780437 17 windows.mouselike.org response in 2ms
1746780438 18 windows.mouselike.org response in 2ms
1746780439 19 windows.mouselike.org response in 2ms
1746780440 20 windows.mouselike.org response in 2ms
1746780441 21 windows.mouselike.org response in 3ms
1746780442 22 windows.mouselike.org response in 2ms

God knows what HO is doing!

Edited by Pippo123456 (Fri 09-May-25 10:07:37)

Yeah I can see it’s a clear BQM from the Firebrick at AAISP…not so from the f8lure server at Zare

https://f8lure.mouselike.org/index.asp?DoAction=View...

In reply to a post by billford:

I'm no expert so doesn't tell me much, except to wonder what happened to steps 4 to 7?
Anyone else?

This is standard practice for some ISP networks. It's not hugely ideal but not per se a problem.

So what’s the score with your connection @Pippo123456 ?

I can see from f8lure @ Second Opinion hosted at AAISP using a FireBrick - that there was clear connection for around 2 hours this morning. Then so back to a wall of red.

Are Hyper on the case now helping you with your fixed IP ping test diagnosis?

Thanks for following -up.

Hyperoptic: they sent an engineer on Monday who verified that the connection is fine and the problem is upstream. In a follow up email they said that they were going to conduct essential maintenance work on my line next Thursday and this would result in a loss of connectivity for 6 hours ."This work will be carried out remotely, so there’s nothing for you to do and we won’t need to visit your property." Hopefully this will solve the issue. I will update this thread with the result.

Broadband test: the clear connection you saw this morning was when the VPN got disconnected. Funnily enough without the VPN the testing updates fine @ Second Option but not at BQM or the "first option test" at f8lure. Also the reverse test with f8lure, that I understand is not forwarded to Second Opinion, always yields a wall of red. I have no explanation for either.

The reason why the Second Opinion test and external ping don't work with the VPN on is that the router tunnels the reply through the VPN as I am not knowledgeable enough with routing to avoid this. Same for me not being able to SSH into the router by excluding ssh on the WAN interface from tunnelling through the VPN

Its no consolation - but I had the same email notice from Hyperoptic about line maintenance that came through yesterday:

Keeping your Hyperoptic service hyperfast and ultra-reliable is our number one priority.
That’s why we continuously monitor our network to ensure you always get the best possible connection.
As part of this, we wanted to let you know that we’re planning some essential maintenance work at xxxxx on 22/05/2025, between 08:00 - 17:00.

Within this time we estimate the activity will result in a loss of connectivity for 6 hours .

This work will be carried out remotely, so there’s nothing for you to do and we won’t need to visit your property.
Once finished, your service will be restored automatically. Should you have any trouble getting back online, we recommend turning your router off for 10 seconds and then back on again.

We’re sorry for any inconvenience caused. We know downtime can be frustrating and will do our best to complete the work as quickly as possible. If you have any questions, please contact us here, email us at [email protected] or give us a call on 0333 332 11 23.

Thanks,
Igor Sabljak
Head of Customer Service Operations

It's not my main circuit (it's my backup) so I'm not particular fussed. I've also noticed that since Monday the BQM graphs for my Hyperoptic connection the spiky minimal evening peak congestion (evident after 8pm) has now vanished and so to (more concerning) the associated small packet loss at this time.

None of these bear any relation to your issue, but mentioning them in the round.

At the end of the 6h maintainance work I am not sure I can notice any improvement with the VPN disconnected. Still dropouts and wall of red on BQM and F8lure despite connecting the Openwrt router directly to the ONT. As usual Firebrick works but it's difficult to read on its short vertical scale.

Repeated speed tests on TB report an F for bufferbloat. I have activated SQM on the OpenWrt, but I am not sure it makes any difference.

In reply to a post by Pippo123456:

At the end of the 6h maintainance work I am not sure I can notice any improvement with the VPN disconnected. Still dropouts and wall of red on BQM and F8lure despite connecting the Openwrt router directly to the ONT. As usual Firebrick works but it's difficult to read on its short vertical scale.

Repeated speed tests on TB report an F for bufferbloat. I have activated SQM on the OpenWrt, but I am not sure it makes any difference.

Have you asked them to allocate you another static IP address. Perhaps from a different range?

Before doing that, could you advise about a router known to work with HO and BQM. My raspberry/openwrt not only does not work with BQM but does not reconnect automatically if the connection drops. That is why yesterday I had to revert to using the VPN….

Are you doubting your router and/or your setup?

I'm running the Hyper connection on a MikroTik CCR, alongside another provider connection.

MikroTik's can be somewhat of an acquired taste, but there'e stacks of routers in all range of price and capability that will manage a BQM just fine. Just one recent (new) example I've been quietly impressed with the new Ubiquiti Cloud Gateway Fiber - which offers great bang for buck in a small form factor.

Hi,

I was looking for a router that provided enough information about the connection (e.g. connection drops) in case I needed to interact more with HO. I have settled on the Ubiquiti Unify UCG. It did not change anything in terms of connection quality. You can see it in action on the Firebrick from Sunday at 16.30 to today at 13.30. The Saturday trace is for the Asus RT-AX88U Pro. Not worth the extra cost given that I am otherwise covered for wifi. Still wall of red on BQM and f8lure. I am willing to bet that HO is blocking them for residential customers.

As you suggested, I have also asked for a different static IP which became active today at 13.30. Range 89.x.x.x. I can PM you the full address if interested. Still solid red other than on the Firebrick. Unusually, the Firebrick trace is squeaky clean. I am not sure it is registering anything

In general, though, I have the impression that changing the IP block has solved the problem. I will keep monitoring and update this thread if not. Thanks a lot for all your help!

For laughs, I have also received (in an email two days ago!) the ticket update on HO sending the engineer on 13 May. "I want to advise you that our team has run a diagnostics and reviewed your connection and advised us that the smaller box that you have is the cause of the issue you have experienced."... As you might remember, the engineer measured the signal at the ONT and concluded that it was not to blame. Their customer support is very kind but a bit all over the place...

Edited by Pippo123456 (Tue 27-May-25 18:42:06)

Alas, nothing has changed. See this post.

Is your Ubiquiti Cloud Gateway set up correctly? It needs to be configured to accept and respond to pings on both v4 & v6.

Yes. Both v4 and v6. Thanks!

In reply to a post by Pippo123456:

Alas, nothing has changed. See this post.

Yeah I don't really understand what's happening with your connection. As said I have a Hyperoptic connection on a business account with a static IPv4 address that literally hasn't skipped a beat.

I'm really at a loss now as to what the root cause of your issue could be. Given you've swapped routers, had them move you to a different IPv4 subnet.

You could try with your routers IPv6 address and see if that works.

Just in case this sheds any light. The first traceroute below is from a machine with a wired connection straight into the UCG (no VPN), whose LAN IP is 192.168.0.1, triggered by a spike (30.5%) in the packet loss. Whenever this happens the (private) IP 172.17.45.118 shows up in the traceroute. This is never the case for a period traceroute.

2025-06-03 09:02:51 TRIGGERED
+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+
| # | Hostname | Loss% | Snt | Last | Avg | Best | Wrst |
+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+
| 1 | 192.168.0.1 | 35.0 | 100 | 0.6 | 1.1 | 0.6 | 1.4 |
| 2 | ??? | 100.0 | 100 | - | - | - | - |
| 3 | ??? | 100.0 | 100 | - | - | - | - |
| 4 | ??? | 100.0 | 100 | - | - | - | - |
| 5 | 172.17.45.118 | 96.0 | 100 | 1.7 | 2.0 | 1.7 | 2.1 |
| 6 | ??? | 100.0 | 100 | - | - | - | - |
| 7 | ??? | 100.0 | 100 | - | - | - | - |
| 8 | ??? | 100.0 | 100 | - | - | - | - |
| 9 | ??? | 100.0 | 100 | - | - | - | - |
| 10 | ??? | 100.0 | 100 | - | - | - | - |
| 11 | ec2-13-39-234-20.eu-west-(13.39.234.20) | - | 100 | 10.3 | 10.4 | 9.7 | 11.0 |
+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+


This second traceroute has run roughly at the same time from a machine with a wired connection to the Openwrt router running wireguard and reaching the internet through the UCG. The openwrt router lan ip is 192.168.1.254, 10.2.0.1 is the IP of the wireguard interface (actually it should be 10.2.0.2 with 10,2.0.1 the address for DNS) , 149,40.48.124 is the VPN server IP address. All trace routes from the connection through the VPN look the same and the monitor never runs a triggered one (reported packet loss in the monitor on this interface at the time was 0%). Honestly I cannot see any difference between the two.


2025-06-03 09:00:13 PERIODIC

+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+
| # | Hostname | Loss% | Snt | Last | Avg | Best | Wrst |
+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+
| 1 | _gateway (192.168.1.254) | - | 10 | 0.6 | 0.7 | 0.5 | 0.8 |
| 2 | 10.2.0.1 | 20.0 | 10 | 2.3 | 2.3 | 2.2 | 2.4 |
| 3 | unn-149-40-48-124 (149.40.48.124) | 90.0 | 10 | 2.4 | 2.4 | 2.4 | 2.4 |
| 4 | vl221.lon-eq7-core-1.cdn77.com (79.127.192.222) | 20.0 | 10 | 3.7 | 2.9 | 2.4 | 4.7 |
| 5 | ??? | 100.0 | 10 | - | - | - | - |
| 6 | ??? | 100.0 | 10 | - | - | - | - |
| 7 | 52.46.166.146 | 30.0 | 10 | 4.0 | 10.7 | 3.9 | 50.7 |
| 8 | ??? | 100.0 | 10 | - | - | - | - |
| 9 | ??? | 100.0 | 10 | - | - | - | - |
| 10 | ??? | 100.0 | 10 | - | - | - | - |
| 11 | ??? | 100.0 | 10 | - | - | - | - |
| 12 | ??? | 100.0 | 10 | - | - | - | - |
| 13 | ??? | 100.0 | 10 | - | - | - | - |
| 14 | ec2-13-39-234-20.eu-west (13.39.234.20) | 90.0 | 10 | 11.4 | 11.4 | 11.4 | 11.4 |
+----+-----------------------------------------------------------------------+-------+-----+------+------+------+------+

Thanks for all your help and apologies for the poor formatting!

I also got a new email of work in the area on 17/06 after writing to [email protected], although the timing may be only coincidental.

Edited by Pippo123456 (Tue 03-Jun-25 11:35:02)

Success!

TBM here

It finally worked with an IPV6 address. It took me a while to figure out how to do it as the HO router does not seem to work with IPV6. But once I got the relevant info from HO and used it with the UCG it all kicked in.

Many thanks for all the hand holding! Still need to figure the source of the dropouts without the VPN, but that is a trickier story....