Back on static IPs because I need them. Performance is still balls.
Anyone know the GRE endpoints? I presume there's a DNS round-robin being used to split my sessions between Midlands and Mancunia as and when opened.
This Hitron is horrible. Going to shove it back in bridge mode and use my EdgeRouter for GRE. Then I appreciate the true 'quality' of the Benu equipment purchased to boost the value of Liberty Global's investment.
Speed. Zoom Zoom. Really, VM? Seriously. This is 'Ultrafast Voom Fibre 3' at 60 quid a month + VAT, offering, err 'Up to' 200Mb as you couldn't be bothered with installing the capacity to allow the 'proper' 350Mb to be ordered, failing as you did on downstream capacity, then upstream capacity, and now it's actually looking like both are under pressure, though I can't tell as, being on static IPs, the performance of that via the Hitron and the Benu Notworks kit is so bad it can cover up, or exacerbate, performance issues elsewhere.
I've had enough of this. I honestly thought some fixes had been done that would have remedied the static IP performance so that I could, finally, have the service I ordered.
Then it became clear that the budget to build to homes, Lightning, is a very different from the build to keep existing customers, even in Lightning networks, running smoothly. After spending nearly a thousand pounds to pass my property and averaging out across the nodal area at probably a similar figure they've combined us with 2 other optical nodes to save on line card ports, and hence capacity. They could fix this in moments by decombining the nodes so that this one and the TWO others get three times the capacity per customer, but presumably that would be a waste and can't have that.
I can't work like this. My WebEx sessions were freezing up due to the extreme inconsistency in delivery of packets, my voice and video conferencing were pixellated in the case of the video and stuttery in the case of both.
This is absolutely unacceptable. VM Business continue to sell a service that they know they can't provide at the standards they claim they will. Adding that to local network congestion, because on VM, every scrap of capacity unused is Liberty Global's return on investment being lowered, causes a poor performance to become an atrocious one.
People are still selling door to door, signing people up, and a small, new estate of a few dozen properties is going onto this same node.
I have politely demanded that they either give me what I paid for, a service of up to 200Mb with 5 static IP addresses, that runs at what would be considered a reasonable rate and is fit for purpose, or come and collect their equipment.
This is absolutely ridiculous. As soon as the network is built and the Project Lightning money is no longer there that's about it for investment. As little as possible for as long as possible.
This local network has 450+ MHz sitting there doing nothing - they could shove some more SC-QAMs in there to relieve downstream, followed by maybe 2 x 192 MHz OFDM carriers in there right now, and fill the upstream spectrum up near 85 MHz with additional SC-QAMs bonding 6+ upstreams. The first channel is at 25.8 MHz, give it 7MHz per channel for capacity and guard band, 8 channels requires 56 MHz. Could run a channel every 7 MHz from 25.8 and just sneak in with the end of the highest channel ending at 85 MHz. If rolloff is a problem near there drop the lowest channel to, say, 20.8 or 22.8 MHz and job done - highest channel now ends at 80-82 MHz.
That's what Project Lighning HFC is capable of. What's it doing? Bonding 4 channels, same as equipment installed 20 years ago, and the forward / downstream path goes up 750 MHz in terms of what's in use, much as the 20 year old stuff does.
They spend a million installing new network to the estate, combine a bunch of nodes together to completely remove any point of having smaller nodes, and run standard frequency plans as if it's the older network, then stick it on the list for resegmentation. Or, to be more accurate, decombination. Literally just no longer combining the nodes together but putting them on their own ports and own bandwidth.... as they should have been all the time, rather than penny pinching to save a few quid on line card ports and downstream channel licenses.
This is an absurd situation. Virgin Media should be ashamed of themselves for not even bothering to try to deliver adequate capacity to new build areas. They have absolutely no excuse not to, beyond that they just don't want to pay for it.
This mirrors exactly my friends issues on installation of Business Broadband from VM, that I helped him with.
Without wanting to detract from any of your points which are well made and certainly relevant to the discussion, are we missing a fundamental issue?
On a different node, with no capacity issues and a different route (Im across a stretch of water from you Ignition) we see EXACTLY the same issues concerning speed on the GRE tunnel. So whilst local capacity is certainly an issue in your case, given that its not here and yet speeds remain the same as your off peak, theres SOMETHING limiting speeds.
The only question in my mind is "Is it the router or the infrastructure?" Given the router will cheerfully do 350 with GRE off, it leans toward a capacity issue at t'other end of the tunnel...
As a Very Insightful Person, I'm here to share my knowledge. I don't work for Virgin Media.