Broadband ISP BT Sees Record UK Internet Traffic of 30.1Tbps UPDATE2
Telecoms giant BT has revealed that the overall volume of internet traffic across their fixed broadband ISP network reached a new peak of 30.1Tbps (Terabits per second) last night at around 9pm, which is up from around 28Tbps last year and 25Tbps in 2021.
As usual the surge in activity above the normal level was driven by a combination of things, such as the fact that six Premier League Football games were all being streamed live via Amazon’s (Prime) video service and another major update was released for the popular Call of Duty Modern Warfare video game (actually it was an update for both Call of Duty: Modern Warfare III and Call of Duty: Warzone).
On Sunday, BT also saw the launch of Fortnite Chapter 5 (video game update), which reflected a 15-25GB (GigaByte) download available across platforms (PC, PlayStation, XBox etc.). By 10am BT was shipping 8Tb/s+ more across the network than a typical Sunday.
Advertisement
A similar surge was witnessed by other ISPs and can also be seen by viewing yesterday’s aggregated traffic data from the London Internet Exchange (LINX), which handles a key chunk of UK and global data traffic through their switches via around a thousand members (broadband ISPs, mobile operators etc.). LINX peaked at 9.23Tbps around 8:45pm (up from a peak of 7.87Tbps in Feb 2023).
Broadband and mobile providers use sophisticated Content Delivery Networks (CDN) and systems to help manage the load from big online events, which caches popular content closer in the network to end-users (i.e. improves performance without adding network strain). This in turn lowers the provider’s impact on external links and helps to keep costs down.
Nevertheless, demand for data is constantly rising and broadband connections are forever getting faster, thus new peaks of usage are being set all the time by every ISP. Ofcom’s Connected Nations 2022 study noted that the average monthly data volume per household on fixed broadband connections increased over the past year to 482 GigaBytes (up by just 6.4%).
Advertisement
UPDATE 1:05pm
Openreach just informed us that they also saw “the biggest ever hourly peak for the data being used during the 8-9pm period on our network in the UK” – yesterday, during the same 8-9pm period. A spokesperson for the operator said: “More than 29.5 Petabytes (PB) of data passed through our broadband network. Typically we see an average of 21 Petabytes (PB) of data during this period each day.”
UPDATE 9th Dec 2023
Full fibre ISP Fibrus also reported traffic peaking 33% above normal for this period. The ISP hit a peak of 299Gbps at 8.50pm on Wednesday (a new record), which is up from their previous highest day – when the new Call of Duty game was released on November 9th (241Gbps).
Advertisement
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on X (Twitter), Mastodon, Facebook, BlueSky, Threads.net and Linkedin.
« Opensignal Rate UK Broadband ISPs by Quality and Speed
Sky UK Puts Sky Mobile and Sky Broadband Staff into One Team »
Hey Mark,
I assume those numbers include EE and all the Plusnet customers that are now using the BTW Broadband Complete package and/or using BT’s transit and peering so can’t be compared directly with their earlier numbers as Plusnet and EE punters moved?
This is traffic measured on the BT 21CN Core Network, which has been shared by EE Broadband and Plusnet Broadband for years.
So it’s the BT Wholesale network?
So that’s why my ping skyrocketed to 900ms and I couldn’t even load websites!! BTs mediocre core network strikes again.
You should see my Vodafone FTTP thinkbroadband graph from last night 160ms pings from about 8pm to 9:30pm last night.
well no, because if it was the core it’d affect everyone, and it didn’t.
Core problem would only affect the folks downstream of it.
To be fair there’s nothing mediocre about the BT core.
Localised congestion does happen at times like these, mind. Last time I mentioned that to someone working at BT they told me I was nonsense and blocked me.
Meanwhile AWS saw localised quality issues on streams to BT/EE customers but, hey, whatever
“We don’t need more than 100Mbps”
“What do you need all that bandwidth for?”
“Why do you want such fast internet”
“Don’t need faster than FTTC”
Well, here’s your answer.
Seems like it is even worse tonight.
At one stage my normally stable 44Mbps was at 1Mbps down and 5Mbps up.
Irony being, Openreach finally lit up the fibres in my part of the village today…we’ll half of it anyway. Not my house unfortunately.
Btw, does anyone want to load up the LINX Wales tab and explain what the hell has been going on since 03/12?
Sure: the usage is so low that it takes very little traffic to make the graph look jumpy.
That looks like a business ISP and is probably a single customer’s data going across the LAN there during the working week.
When the baseline is a hundred Mbit or so a single gigabit business customer will make a substantial dent.
If my ISP peered with another with a meaty enough connection I could make the previous days’ graphs look like noise. A bunch of folks on gigabit and especially multigig could make it look noisier and the earlier stuff flatter. All about the baseline.
You’ll probably see, as with last week, lower usage 9th and 10th, Saturday and Sunday, then the same bursty pattern throughout the working week. No novelty to it: normal on an exchange with so few flows from so few customers going across it: not enough traffic from other customers to smooth out that one customer’s bursts.
I rather see Altnets publishing their data.
You aren’t missing much. Usage per customer is likely a bit higher than Openreach as no xDSL but altnets with any scale track usage on the Openreach network pretty closely as everyone has the same content available.