Decided to bring home a fibre cleaning kit and cleaned the ONT end fibre, reseated and power cycled the router, things had a glimpse of looking better!

After messing with some WiFi settings as the prior engineer reset the hub, the 5X completely bugged out, and stopped handing out DHCP, wired connection also wouldn’t obtain an IP, and a static IP also resulted in no ping to 192.168.0.1.

Rebooted the ONT/router again, the LED light didn’t even change on the 5X and stayed solid white even straight after powering on, and had the same issue.

Rebooted for a third time and reseated all ports including the fibre which managed to get me back online and DHCP started working again, however back at 2100mbps download and 12mbps upload.

I think this is an ONT/router fault, I can’t explain that behaviour from the 5X. Going to ring them later and explain and hope they actually understand, if they tell me to factory reset the router for the 5th time I might spontaneously combust.

In reply to a post by suttonscloud:

I think this is an ONT/router fault, I can’t explain that behaviour from the 5X. Going to ring them later and explain and hope they actually understand, if they tell me to factory reset the router for the 5th time I might spontaneously combust.

Yes definitely doesn’t sound healthy. Needs replacing. Good luck with the service team there.

In reply to a post by suttonscloud:

Just wanted to see your thoughts and if anyone else experienced this on XGSPON. I obviously know the ins and outs of fibre networks with the nature of my business but just wanted to see if it was usual behaviour or potentially even a overlook on my behalf and the connection was fine and it was the BQM requests that were being dropped by the router.

Don't worry about the BQM for now: the fact that your upload speed is so terrible is what matters.

I think the likelihood is that the poor upload speed is caused by high packet loss upstream, which in turn may mean the optical signal is not strong enough in that direction, but it's not your job to diagnose this.

Aside: measuring packet loss separately in the inbound and outbound directions *can* be done with suitable software at both sides of the link. The perfsonar toolkit is one example, although that's a large hammer to crack this particular nut. The underlying tool is owamp for those measurements, so you could just run owamp on a local machine and on a cloud VM somewhere.

By default perfsonar sends 10 packets per second in each direction; this means that over an hour you can measure packet loss as low as 1/36000 = 0.003%

In reply to a post by candlerb:

I think the likelihood is that the poor upload speed is caused by high packet loss upstream, which in turn may mean the optical signal is not strong enough in that direction, but it's not your job to diagnose this.

The extreme packet loss and poor upload are directly correlated, almost without shadow of doubt. Up to VM to get to the bottom of it.

In situations like this a resi-based service can be a nightmare to get through to the team that can actually diagnose and help, in a timely manner. I feel the OPs frustration here.

In reply to a post by suttonscloud:

Strongly disagree with this comment! - VM Community forums have been a nightmare with me, i have had nothing from them, actually raised a complaint from dealing with them. The Forum team wont deal with XGSPON customers, they outright refuse.

That is a shame, I guess because Nexfibre run the network they have no visibility. As someone in an older DOCSIS Coax area I have had all sorts of issues the call centre were useless but the Community Forum was amazing.

I guess it means VM ISP via Nexfibre is not really VM... and the two companies need to get their joint act together. The sooner Nexfibre offers other ISPs the better for you and my colleagues.

In reply to a post by candlerb:

In reply to a post by suttonscloud:

Just wanted to see your thoughts and if anyone else experienced this on XGSPON. I obviously know the ins and outs of fibre networks with the nature of my business but just wanted to see if it was usual behaviour or potentially even a overlook on my behalf and the connection was fine and it was the BQM requests that were being dropped by the router.

Don't worry about the BQM for now: the fact that your upload speed is so terrible is what matters.

I think the likelihood is that the poor upload speed is caused by high packet loss upstream, which in turn may mean the optical signal is not strong enough in that direction, but it's not your job to diagnose this.

Aside: measuring packet loss separately in the inbound and outbound directions *can* be done with suitable software at both sides of the link. The perfsonar toolkit is one example, although that's a large hammer to crack this particular nut. The underlying tool is owamp for those measurements, so you could just run owamp on a local machine and on a cloud VM somewhere.

By default perfsonar sends 10 packets per second in each direction; this means that over an hour you can measure packet loss as low as 1/36000 = 0.003%

Thanks @candlerb,

I will look into this tonight as i do have a AWS farm i can spin up an instance there, and a DL380 G10 at home in the lab, so ill spin up two VM's on there and see if i can capture anything, more for my own personal gain rather than trying to prove a point to VM.

Is there anybody else close to you with VM? If they're close enough to be on the same PON then you could swap Hubs with them to rule yours in/out as the issue.

Another option might be to take your Hub outside and plug it directly into the cable coming from the street to rule out an issue with the internal lead in.

In reply to a post by Pheasant:

In reply to a post by candlerb:

I think the likelihood is that the poor upload speed is caused by high packet loss upstream, which in turn may mean the optical signal is not strong enough in that direction, but it's not your job to diagnose this.

The extreme packet loss and poor upload are directly correlated, almost without shadow of doubt. Up to VM to get to the bottom of it.

In situations like this a resi-based service can be a nightmare to get through to the team that can actually diagnose and help, in a timely manner. I feel the OPs frustration here.

So this morning, i'd had enough. Thrown the 5X in the bin and now have used my own custom SFP with the 8311 software, configured my PON address, S/N and MAC address and managed to get a link, put it into my OpnSense server and voila, things are magically better.

Rubbish 5X router looks to be the problem.

I have enabled ICMP on OpnSense so the BQM works, and changed the ip so 'fingers crossed' the graph from 15/01 09:30 should look a LOT better.

Getting 5ms ping responses from most places, facebook.com, google.co.uk, etc etc.

Speedtest here

Back to sleep for another night shift later.....

Good news. 👍

In reply to a post by Pheasant:

Good news. 👍

BQM looks better than ever today, so looks like definitely a faulty router.

Only had a few tiny spikes of packet loss, which was probably me running speed tests.

Long live the WAS-110 SFP.