
The London Internet Exchange (LINX), which handles a key chunk of UK and global data traffic through their switches via 950+ members (broadband and mobile operators etc.), has today announced that alternative full fibre network provider toob has become the first ISP to take a 400G (Gbps) port at the IXP for their peering services and more.
Just for some context. Toob is a Hampshire-based operator that was originally backed by £75m from the Amber Infrastructure Group (here) and “up to” £87.5m from the Sequoia Economic Infrastructure Income Fund (here). During 2023 the operator also secured £160m of additional funding (debt financing) from Ares Management‘s Infrastructure Debt strategy (here), which could be upsized to £300m over time to support growth.
Toob’s own FTTP broadband network is known to cover 150,000 premises (24th Aug 2023 – not all RFS) and they’re aiming to reach 300,000 premises across parts of Dorset, Hampshire, Surrey and Sussex in the future. In June 2024 the provider also revealed that they’d passed 50,000 customers (here), which is more than double the 20,000 they had a year earlier.
Advertisement
Suffice to say that they’ve seen a fair bit of growth and as part of adapting to that they’ve taken the first 400G port at LINX, which actually introduced 400G ports towards the end of 2021 due to customer demand. LINX enlisted technical supplier Nokia to deliver this additional service. But so far at LINX’s UK platforms, it’s mainly global content delivery (CDN) networks or global carriers who have traditionally had demand for this solution.
Sean Teggart, Technology Director of toob, said:
“We are thrilled to pioneer the adoption of 400G technology at LINX, reaffirming our dedication to delivering unmatched broadband connectivity in the UK. This strategic upgrade not only enhances our network’s capacity but also increases toob’s reliability and prepares us to support the future needs of our customers.”
Jennifer Holmes, CCO for LINX, said:
“We are really excited to be supporting and celebrating toob during their upgrade to 400G, the first UK ISP to take this step. This reflects the continued growth and demand of online services, good connectivity and the Internet as a whole in the UK.”
The new 400G port appears to have been connected within LINX’s largest LON1 (London) Ethernet switching platform, which is accessible from 16 data centre locations.
Toob’s investment is welcome but misplaced. They’re still a fledgling ISP, with a lot of problems to solve. Only last week they had a mass outage (Hampshire area, maybe further afield too) that took out all services – residential, business, even their own website and phone system. Their support teams lack knowledge and foster unnecessary expense too – they’d sooner send an engineer out with a new router than tell you how to reset yours.
How full was their existing LINX port?
Marketting fluff from Toob+LINX. They aren’t hitting anywhere near the limits of their port, but with the LINX port pricing they could take a 400Gbps port, but have only 100Gbps of ‘peering service’ on it.
Lots of ways to make yourself look bigger than you are, not really sure eyeball ISPs always want to do this.
Will they be upgrading their LONAP capacity now too?
@Anon you don’t know what you are talking about.
Scaling my org, With 50K customers they can easily be doing 70G peak over LINX. 70% is a fairly standard upgrade point. The LINX port pricing makes 400G the next logical step rather than 2x100G if you have the ports available.
All press releases are marketing fluff, but this one seems like a sensible upgrade, and an interesting note that the large eyeballs who peer at LINX haven’t yet had to make that jump.
@Anon they already have 400G of capacity at LONAP — I don’t think it needs further upgrading?
Single pint of failure?
No. Will take transit if the LINX port fails.
Or shift traffic to LONAP. I would expect toob to have modelled what happens if their LINX / LONAP / etc. connection failed, and verified that they have enough headroom on alternative connections to be able to handle the additional traffic.
With their customer count they wouldn’t be close to maxing out either port even if it took every single bit their customers transfer indeed.
Their customers probably peak, rarely, in the ballpark of 50-60G and sit mostly at 30-40.
@IX you are wildly lowballing. The reality will be close to 200G for 50k customers.
Relying on LINX rather than having your own direct peering isn’t exactly something to shout about.
They’re almost certainly peering publicly over LINX as well as privately — although given how datacentres are price gouging cross connects I wouldn’t say it’s unreasonable of toob to prefer a few large public peering connections over several 10G / 100G XCs.
Having private peering that hardly passes any data because someone thinks using an IX LAN is somehow inferior is dumb.
If they’ve 200 peers on LINX that means 200 switch ports. Or just use one. It breaks, you have an SLA with the IX.
They’ll have PNIs to key, high throughput peers where it makes sense for both alongside maybe on-net CDNs.
Have plenty of peers available as can be seen here: https://bgp.he.net/AS60377#_peers
SkyB uses 500G port on Linx1 according to PeeringDB
News to LINX. Sky have 100G to each LAN according to them.
I don’t think the LINX member page deals with LAGs properly. I’ve had look at a few others that have more than 100G to the LANs and they all report 100G.
Sky will have a 5x100G link. A 500G port doesn’t exist.