ISP KCOM Grows UK Broadband Customer Base to 139,700
KCOM, which recently extended their Fibre-to-the-Premises (FTTP) network to cover 250,000 premises across East Yorkshire and North Lincolnshire in England (here), have published their annual accounts and revealed that their total broadband customer base has grown to 139,700 (most are now on fibre, having migrated from copper).
The operator had previously invested £85m+ to spread FTTP across their core network patch of c.195,000 premises in the Hull city area. Soon after that the operator was acquired by Macquarie Infrastructure for the hefty sum of £627m (here) and in early 2020 they began a large £100m network expansion into more competitive areas (here), which continues today.
At the last count, in June 2021, we estimated (here) that KCOM were currently in the process of adding around 79,000 extra full fibre premises to their existing coverage (i.e. since the expansion programme began in 2020). So far, some 20 additional towns and villages across East Yorkshire and North Lincolnshire have benefitted from this work.
According to the company’s latest annual accounts, KCOM added a total of 26,000 new premises passed to their network coverage in the last year alone and generated total revenue of £99.6m for 2021 (up from £99.2m in 2020), with profits before taxation of £4.3m (up from -£14.3m).
“The expansion of our full fibre network to surrounding areas ramped up during the year and as at 31st March 2021 we had passed 28,700 premises and we are on track to complete the build in the second half of 2022,” said KCOM’s report. The 2022 completion target represents their Phase One expansion, and there’s talk of planning work being undertaken for a second phase, which may partly focus on smaller “rural locations“.
The operator has a total customer base of 147,000 (up from 141.1k), although these aren’t all broadband users. When split down, KCOM had a total of 132,300 consumer broadband customers (up from 124,400), with 117,800 on FTTP. On top of that, they also had a total of 7,400 business fibre broadband sites (up from 6,600). The overall broadband base is thus 139,700, with 229,000 premises passed by FTTP, as at 31st March 2021.
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 and Linkedin.
« Vodafone UK and EE Delay Reintroduction of EU Roaming Charges
Latest UK ISP News
- FTTP (5533)
- BT (3518)
- Politics (2542)
- Openreach (2299)
- Business (2267)
- Building Digital UK (2247)
- FTTC (2045)
- Mobile Broadband (1978)
- Statistics (1790)
- 4G (1669)
- Virgin Media (1621)
- Ofcom Regulation (1466)
- Fibre Optic (1396)
- Wireless Internet (1392)
- FTTH (1382)
The fibre service having been a fibre customer for the last 6 years (and a 56k & ADSL customer for around 20 years), has been shoddy and terrible since September 2021 as a result of this (half price to new customers and spreading out to surrounding areas, agressive traffic shaping of gaming traffic from 6pm till midnight limiting console gaming to around 10.1mbps.
It seems to happen most nights of the week if it happens everyday that’s 2190 hours a year just over 3 months if a 12 month contract where upload speeds are shaped that low packet loss occurs rendering online gaming pointless.
I couldn’t be more disappointed with my “future proof” super fast “perfect for gamers” service just lately. This is not true unlimited internet… It’s very limited more or less 6 hours a night…. Every night
I’ve noticed the same issues for the last couple of months, 6PM to Midnight or shortly after, it’s mostly rendered my streaming of BT Sport (Ultimate/4K) for the Champions League (8-10pm) impossible and buffers or frequently drops to lower quality even on the 1080p version.
I’ve not had this kind of experience of the service at all in the similar number of years I’ve had FTTP it’s previously been full speed everywhere all the time.
I’ve found a interesting network tester at http://cloudharmony.com/speedtest-for-limelight:cdn-akamai:cdn-fastly:cdn-and-aws:cloudfront (only bothered about the 256KB – 10MB test)
They all slow down to 5-50Mbps (or worse) at peak times (all tested wired).
By 00:30 back up to >400Mbps on all services.
BT Sport seem to use LimeLight and Akamai CDNs, they even seem to use internal KCOM network caches but still throttle them (???).
Also noticed the same with Nvidia downloads (Edgecast CDN) full speed all day (>40MB but 1-3MB at peak)
I’m not really sure how to go about complaining to KCOM, I don’t fancy ringing up to be asked if I’ve tried restarting the router since the problem is obviously deeper within their network caused by congestion or traffic policies.
I think I’ll try emailing them eventually.
I’ve been on the phone to tech support numerous times, even had a call back completely useless experience, she was trying to convince me it was Sony servers throttling me and would not entertain the fact that they’re traffic shaping,
I had one tech support guy tell me they were throttling my speed about two years ago, it took me over 8 hours to make a dent in a 47gb cod warzone update he told me because it was so popular they were shaping traffic to lighten the load. He assured me it wouldn’t affect the online play side of things just downloads, after he reset my profile and I reset my router it came down at full speed.
The lady who called me back last week bluntly denied that was happening and had me run a speed test… Baring in mind this speed test was at 4pm the day after. It obviously came back correct and gave her the old everything is normal it’s Sony or your hardware, I again tried to get my point across that it’s protocol shaping that’s taking place so even during peak times there is a massive chance a browser speed test will report advertised speeds anyway as it’s not being shaped/filtered, she was unwilling to see my side of it continued to treat me as some dummy repeating the same thing to me for around the 5/6th time … I got sick of going round in circles and ended the call (win for kcom I suppose).
That night about 3hours later, back to a strangled speed.
I went digging around on the website and come across there “traffic management information”, it clearly says email/streaming/p2p/ are not filtered (there is a footnote stating they will if they see fit though, so probably do. Out of all the protocols the only one they do filter is gaming downloads and gaming updates even though they sell packages “great for gamers” I could accept this to some degree and limit my downloads until after midnight but they’re limiting my upload to an average of 10.1mbps, theta like 1.25MB/s worse then my old ADSL line. That’s got to be against my ‘consumer rights 2015’ as it’s not fit for purpose.
I’m looking at putting in a formal complaint, then I’m planning on contacting the ADR -alternate disput resolution and the onbudsman finally citizens advice.
It’s a complete headache but something has to be done as there is no way I’m paying above average price for speeds I’m likely to be frozen out of for 3 months of the year.
I’ve had nothing but issues nearly every night from September it isn’t right at all. After they wrecked my gaming session two nights ago I gave in and loaded Netflix instead, it kept downgrading a 1080p movie 3gbps to 270 at 0.5mbps at that resolution it was like watching a Lego movie filmed on a VHS camcorder.
Why they’re knocking out 900mbps packages is beyond me if they’re throttling everynight the network clearly isn’t upto it or more likely they’re loving getting fat coin off people for a service they don’t have to deliver and saving loads of bandwidth that should be going on customers packages.
The service has gone so far downhill since Karoo/kcom was sold to Canada and I feel it’s only going to get worse.
Hi Kev. I live in one of the villages that is getting KCOM and was very happy to finally be getting fibre broadband, unlike most of the village who are complaining about paths being dug up. I’m very dissapointed after reading your’s and Miken’s posts, I can see there are a few other similar facebook posts too. I also found there FAQ on traffic management, which basically says traffic management is used during peak hours for everything if the network is busy, which im guessing it now is.
Is traffic management used during peak hours? No***
***KCOM does traffic manage during extremely busy periods. This is on a network level.
Even though i only have 44mbs download and and 5mbs upload, it looks like i will be far better of sticking with BT’s fttc as we mostly only use the internet between 6pm and 11pm and mostly for gaming or streaming. At least mine never ever drops quality… or slows down
You say it is like being out of service for 3 months out of the year but I would say it would be more like being out of service the whole year of the year. Seens as the time you say it is bad is the only time I would use it (after work-excluding weekends)
I hope KCOM sort it out soon for you and everyone else.
I think one avenue could be the net neutrality side of things.
Ofcom rules are on https://www.ofcom.org.uk/phones-telecoms-and-internet/advice-for-consumers/advice/net-neutrality it’s easily demonstrable that they are throttling many CDN providers (see cloudharmony).
Interesting you mention Sony issues, a quick google reveals they probably also use Akamai and Limelight, I haven’t setup packet monitoring myself to check.
I do seem to find Limelight suffers the most.
KCOM are Akamai Accelerated Network Partners, they host kit to serve their users, it’s very unlikely they have capacity issues with traffic that is hosted within a few miles of their users and if they did I’m sure Akamai would just add capacity.
I see Akamai downloads hit servers in my local exchange (the whois description has my exchange name) and is 1ms away.
I do agree with them that there could be capacity issues with CDNs at times (after popular game updates), but maybe only for a day or two after release, not ongoing for months.
You also mention Netflix, again similar program to Akamai hits servers probably in my local exchange, I haven’t tested it but it should be able to provide full speed all the time (it probably does since fast.com uses those same internal servers).
I can remote into my work via VPN (KCOM leased line 500/500) and force it to download from the exact same internal Akamai server and get full speed at peak.
I also came across the traffic management policy https://www.kcom.com/home/legal/traffic-management/ which only mentions that gaming traffic could be slowed down “*note Gaming updates and software downloads may be allocated a pre-determined level of bandwidth, network wide, to ensure that high demand for this type of traffic does not impact other traffic types.”
I don’t think they can distinguish game updates and software downloads from other time sensitive traffic (BT Sport streams) since it’s all just encrypted HTTPS traffic so they’ve probably just used a heavy handed traffic policy on all known CDN IP ranges.
I also noticed similar problems with the football on Amazon Prime, it was being served by Limelight at the time.
I might try a VPN service or setup my own with some policy based routing sometime to see if that makes any difference (I may even try routing Akamai to go HOME > [LONDON > KCOM AKAMAI > LONDON] > HOME)
I expect I’ll have problems with accessing somethings though; The AAISP l2TP service could help but with the price I pay KCOM I don’t really want to add that.
I really don’t know how you get past level 1 techs and to someone who can actually investigate, maybe the complaints procedure is the only way.
On some level I understand they need some traffic management, but whatever they’ve done evidently doesn’t work its rendering the connection unusable or affecting other companies services.
Same issues for me too. Gaming is a nightmare at the minute. Packet loss and general slow downloads are making it hard to live with. £55 a month at the minute for a 400/80, which goes to complete garbage after 5/6pm.
Complain on twitter/email etc. Something needs to be done.
Thanks for the speedtest site, be interesting to test that later on a wired machine to see what happens.
If anyone has any suggestions on how we can get to the bottom of this, feel free to share!
Well I don’t think they are treating traffic differently, everything is just severely limited intentionally or not.
https://www.thinkbroadband.com/speedtest/1641587379811341355
3.3Mbps for a single stream!
Well according to the connexion website .
125 – £37.50
500 – £44.99
1000 – £49.99
18 month contract
Free WiFi 6 Router
Smart phone app
Should be interesting
Like most I look forward to Connexin taking KCOM on the competition can only be good for consumers, but I don’t think they’ve released any rollout plans? based on how long KCOM took it could be many years away for most areas.
I’ve tested with a couple of VPNs this evening, I hosted my own on DigitalOcean and Linode, Linode was the most interesting so I’ve only including those results the other one was similar, surprisingly BT Sport worked via Linode.
No VPN
https://www.thinkbroadband.com/speedtest/1641672906452169255 – single thread 13.2 Mbps
Nvidia 10-50 Mbps (if I’m generous)
BT Sport loading 5MB chunks in up to 5 seconds – basically on the edge of buffering and it wouldn’t be fast enough for UHD (checked and device checker in BT Sport app fails)
Linode – Wireguard VPN
https://www.thinkbroadband.com/speedtest/1641674934562359255 – single thread 241.5 Mbps
Nvidia 200-350 Mbps
BT Sport loading 5MB chunks in 250 Mbps)
Likewise, the cloud harmony tests (Limelight, AWS, Akamai, Fastly) all jump up to 200-400 Mbps instead of 1TB would get expensive).
I’m not sure why but half my message got cut off there.
—
Curiously even without the VPN I can still get near enough full speed to http://speedtest.london.linode.com/100MB-london.bin but not to other things, only thing I can see is outbound goes via LONAP and return seems to be be LONAP or LINX.
I’ve started looking into the KCOM complaints procedure and may put something together for that.
If I was going to use a VPN as a solution (even just for the BT Sport on Fire Stick) I think I’d probably use an actual VPN service, I just used the above because it didn’t cost me anything to try it (>1TB would get expensive).
Good work Miken. We need someone like yourself with a bit of knowledge and technical know how to give them a good run for their money!
Stick in a complain through the complaints procedure. Happy to do the same too!
Apologies for lack of replies, not been well lol, I registered my interest with connexin in December when I started to get most annoyed with my kcom service, you receive a promotion email offering first two months free if I sign up now, on the website it states they can rent a third party line until it’s up and running but the data is very vague and I can’t see how it would work out cheaper if they do indeed rent out the line from kcom. I definitely don’t want a wireless delivery of internet to my home.
I keep meaning to call them to get information on this and there useage/traffic management policy but I’ve just not felt upto it lately. I keep meaning to buy a VPN service as my subscription has currently ended but I don’t fancie extra expense just to use my paid for bandwidth at full speed (in principle).
I am still planning on a formal complaint and a call to connexin (as soon as I feel upto it).
I primarily only posed here and on downdetector to highlight the otherwise hidden policies kcom had in place as I knew kcom was lying to there customer base on a daily basis and as I don’t use the usual social media platforms at all it seemed like the right thing to do.
It is a problem a massive one that needs to be addressed as it’s only likely to get worse.
Hope you are all well 🙂
I’m pretty sure where Connexin isn’t available they just resell KCOM (there isn’t anyone else they could be using), no idea if the traffic gets handed off to Connexin or if it’s just white labelled KCOM packages.
I’ve submitted an essay/complaint via email, its the first one I’ve felt the need to do since 2014.
They aim to respond within 2 working days to acknowledge the complaint but actually resolution will take longer.
I’ll comment here when I hear something.
Nice one Miken.
I’ve had a few issues in the last hour or so with dropped connection. Router reboot sorted it but looks like a few others have had the same thing according to downdetector.
An update on the progress so far.
I received a reply to the email/complaint within 6 hours, just to organise a call to discuss and verify our details.
10/1 20:30, verify the account password, they didn’t need me to do any further testing during the call and would be escalating the issue – so it seems I’d presented sufficient evidence already.
Explained the next step would likely involve placing a Raspberry Pi on the connection so they can do some diagnostics/testing (I didn’t ask what they would do but I’d expect iPerf or hole punching for SSH access).
11/1 13:00, update callback, they are trying to source a Raspberry Pi a different department is involved in that so it could take a couple of days.
Accepted that there does seem to be something limiting our speed.
12/1 14:40, another update callback, the network escalation team have noted a couple of other similar reported issues therefore the need for a Raspberry Pi has been put on hold while that investigation takes place. They’ll update me once they have some information about that investigation.
I must say the response I’ve had so far has been excellent, at no point did they try to dismiss my claims, both employees have understood how I described the issue (I’m glad since I spent a while over the weekend rewriting it to try and make it as clear as possible – it was still 1000 words lol), since the initial escalation the same employee has been calling me back so I haven’t had to re-explain the issue (I haven’t even needed to they understood what I’d wrote).
At no point have I mentioned traffic shaping, I don’t think whatever is causing the issue is intentional since it makes no sense for them to be limiting internal network speeds (several CDN/caches, Akamai, Limelight, Google (YouTube), etc.) likewise putting a VPN on restores full speed.
I did think their use of Raspberry Pi’s for diagnostics was interesting, must be a hell of a lot cheaper than the kit they needed for testing ADSL lines, plus they can leave it at a premises for days/weeks.
I don’t have any idea of timescale, there is an investigation going on and they’ll update me when they can.
Honestly, I don’t mind even if it takes a week, month or longer I’m just glad it appears to be being investigated thoroughly.
I had someone call me today regarding the comments I made on the connection during peak hours. I was told to basically ring up when there was an issue. They’ll check the line, do a speed test and if an engineer comes down and there’s no fault then i’d be charged etc.
I had tried explaining that the issue wasn’t at my end and was something at their end causing issues during peak hours (Packet loss, slow downloads etc)
Said i’d give them a call, but want to be armed with a bit more information, speedtests from thinkbroadband/CDN’s etc during peak hours in comparison to off peak.
Hello Kev, KCOM here.
We’ve read your post and are sorry to hear that you’re having issues while gaming.
We’d very much really like to investigate the issues you’ve described to find a resolution to the problem. We can’t message you directly on here but we’ve provided our email address below – if you’re willing, we’d like you to email us with your details so we can contact you:
question@kcom.com
In the subject write ISP Review so we can easily identify you.
Also, if anyone else on this thread is experiencing similar issues as Kev – Miken and James – please also do feel free to get in touch as we’d love to explore what’s happening to your service further.
We hope to hear for you soon.
Regards,
KCOM.
Hmm.
8:30pm
https://imgur.com/kmAI187
That’s not great is it…
And this too.
https://www.thinkbroadband.com/speedtest/1642106418192031355
Glad to see KCOM keep an eye on the articles 🙂 I did mention comments on here in my report.
I’m seeing the same here at about 10pm, personally I do find 8 to 10/11 is the worst time, then back to normal by sometime around 11 to 12.
Limelight is always the worst affected on CloudHarmony, it seems to use London servers for it rather than KCOM caches which I’ve seen BT Sport use (no idea why, maybe ISPs only cache popular content or something).
The line is otherwise idle (my router shows current usage).
2.1 Mbps on Limelight
https://imgur.com/tPoQQDg
7.3 Mbps, that’s half what we used to get on ADSL, but 6x connections are way faster.
https://www.thinkbroadband.com/speedtest/1642111050646224555
Similar speeds show in Youtube (right click and “Stats for nerds”) or the Netflix debug menu – they both show your current connection speed.
But run a speedtest.net or fast.com test to KCOM servers and it looks all hunky dory.
I’ve fired an email to them today and then another tonight with the above results too
Speedtest.net is my go to and you’re right, all looks fine.
Will see if I get a response tomorrow.
Be interesting to know if people on the 100mb package have the same performance as people on higher packages.
If so, why are we paying £15 more a month, (or more if you’re on the 900mb package)? When 6 hours of the day we’re getting the same speeds as someone on a cheaper package?
Hmm.
@KCOM.
I appreciate you reaching out and I have sent my details to the given email. I look forward to your reply.
Apologies for the delayed response.
(Nice job on the formal complaint Miken)
I hope everyone is doing well !
Hi Kev.
I sent an email over to them Thursday last week but yet to hear a response, other than the automated reply.
Thanks for the heads up James,
I’ve also received the automated email,
I will update the post/thread with any other information/correspondence I get from kcom (or lack of, in the event I hear nothing).
Hope your well, stay safe!
It’s strange that neither of you have heard back, they were pretty prompt with me and to a reply.
It’s been over a week since I last had a callback from them, I just tried ringing them but they couldn’t give me any update.
It’s still as bad as ever, yesterday 3.1 Mbps and 8.3 Mbps
https://www.thinkbroadband.com/speedtest/1642624277400935655
https://www.thinkbroadband.com/speedtest/1642629214489139855
I did notice a speedtest at https://speed.measurementlab.net/ can report what seems a fair amount of retransmissions at times (8%), the test will be TCP hence they are retransmissions but if it was UDP it would be packet loss (things like gaming will be UDP).
This is purely speculation (I don’t know if KCOM use Cisco but I expect whatever they use has similar features) but I was reading https://www.cisco.com/c/en/us/support/docs/quality-of-service-qos/qos-policing/19645-policevsshape.html and the disadvantages of traffic policing sound an awful lot like the symptoms we seem to see.
“Drops excess packets (when configured), throttling TCP window sizes and reducing the overall output rate of affected traffic streams. Overly aggressive burst sizes may lead to excess packet drops and throttle the overall output rate, particularly with TCP-based flows.”
I believe James is but Kev are you also on the 400/80 package?
Just wondering if it could be a misconfiguration with this package, since I don’t think it was long after we were forced to this package that we started having issues.
Hi Miken.
I had a response yesterday or the day before to confirm address etc and then was told documentation I provided was added to my case.
You’re right, I’m on the 400/80. I too had the same thought the other day, a misconfiguration of this package. They pushed these onto us around early september which ties in with the timings
I’ll have a look at the other bits you’ve mentioned in the post tonight and see what figures I get.
I did the CDNs last night and it was horrific 3-8 on all with some failures!
Packet loss on gaming didn’t seem as bad, so not sure if something’s happening in the background?
Hi Miken.
I believe I’m on the 300-D/75-U package.
We was looking at moving up to the 500/100 but there is very little to no incentive in doing this given the aggressive traffic management.
I’ve noticed failed tests on the cloud speedtest during peak hours too.
I’ve noticed I loose 50mbps down and around 20mbps up during peak times not to mention 1/2% jitter and 2/4.0dd% packetloss.
I’ve not been loosing packets when gaming as much over the past few days but I’m still less than happy with my service, fibre was supposed to be the all mighty but it’s so heavily policed I miss my ADSL line.
Kcom never behaved like this now they’ve gone so heavy handed after being a customer for 15/20 years I’m about done. I don know how it can be called a unlimited service, I’d find it funny if it wasn’t so tragic.
I had pure broadband knock on the door offering FTTP today, not to mention connexin hooked up its first fibre customer in my area two days ago.
It’s a real stupid time for Karoo to be giving a terrible customer service so many new companies offering fibre I never thought I’d leave always been happy with the service but now I’m looking to go just out of principle, it’s a lot of money to be starved out of your bandwidth.
My last call with tech support was the straw that broke the camel’s back… Constantly shifting blame elsewhere and not willing to hear the customer out.
I’ve still not heard nothing from them.
Hope you are both well.
Not heard anything back as of yet.
No news here either I’m afraid, I’ve not heard anything for two weeks.
I did notice they have had a couple of people tweeting them complaining about evening speeds.
I just spotted this tweet from a Hull IT MSP https://twitter.com/weareability/status/1486300375015636992
I’ve no idea if it’s the same issue but they are suggesting a speed issue may only affect older ONTs.
Dropped them an email and got a response back fairly quickly.
On about installing monitoring equipment?
I sent them an email this morning asking if they had any update.
We’ve organised getting a Raspberry Pi installed on Wednesday to be recollected Monday next week.
It’s to allow them to monitor and test the connection, as I mentioned before I haven’t asked exactly what they’ll be doing but I’d expect they’ll run something like iperf/iperf3 to basically run speedtests between our end of the connection and somewhere else on the network, maybe reconfigure something and test again.
Hopefully* they’ll be able to recreate the issue (as long as it’s not to their speedtest.net server since that doesn’t seem to be affected!), diagnose what causes it and fingers crossed resolve it.
* I can’t see how they wouldn’t, I’ve been regularly testing using a variety of testers along with normal CDN downloads and the problem persists.
The Raspberry Pi was installed today, a quick observation from it’s traffic shows that every 10 minutes it pings a bunch of places (not limited to twitter, yahoo, ebay, bbc, etc. – it also tries facebook.co.uk which doesn’t exist…) and does a speedtest against fibretest.kcom.com (same as speedtest.net uses)
So I’m almost certain that they aren’t going to be able to recreate the problem since I’ve never seen their own speedtest slow down.
I can see it running the tests https://i.imgur.com/gEq25mT.png (upload/download are flipped) but full expect to see this performing exactly the same all evening.
It’s not getting collected until Monday so its data collection wont be analysed until later next week, then I’ll probably have to try and convince them that their monitoring methodology is flawed.
I might write my own monitoring tool sometime, that picks a non-KCOM Speedtest server.
Jeez.
Just when you think you’re getting somewhere they rely on the one thing that’s not showing an issue!
I Might as well not bother with the PI if they offer if it’s not really going to help, or show the issue.
Hopefully you have look trying to convince them!
Why’s it’s pinging addressed that don’t exist either? Seems bizarre.
I think it’s just been setup with a bunch of domains to ping and they’ve put facebook.co.uk in there, so every 10 minutes it tries and fails to ping it because it can’t resolve an address.
As expected I can see the speedtest is full speed throughout the evening, I tried a bunch of other speedtest.net servers myself and wouldn’t you know it they all are as well.
So I think the measurement device will be active for 5 days and not capture anything that shows the issue.
I think I’ll capture some packets for thinkbroadband, measurement labs and maybe a file download to send them including “normal” ones in the morning, it can’t hurt even if they want to wait for the measurement device results.
Yesterday I sent them some videos/screenshots of YouTube 4K (the stats showed it was struggling to buffer 1 second), Nvidia download, TBB and M-LAB along with some packet captures from the evening of the 2nd.
I haven’t heard anything back about that, I’m not even sure if the email got through since it had a load of links to mp4’s and zips.
I did notice that I didn’t have any issue watching YouTube 4K yesterday evening (3rd).
Tonight (4th) YouTube 4K is working, TBB, M-LAB and Cloud Harmony are all testing much better and some not too far from off-peak performance.
It’s too early to say it’s fixed but maybe something has been happening to resolve the issue I’ll have to see how it is over the weekend.
Speed seems a little slower to ramp up than off-peak but it does eventually hit full speed.
An iperf test to a server in France that off-peak was full speed is still struggling.
I also noticed they’ve posted about essential maintenance over the next couple of weeks on https://www.kcom.com/home/help/ I wonder if that could be related and what “involving development activity” means.
I’ll have a look tonight if I remember. I have noticed less packet loss on gaming lately so maybe something is happening.
Be interesting to see if they get back to you with an answer to the stuff you sent them.
For me it seemed OK Saturday, but Sunday slow speeds and high pings as usual again 🙁 (even kcom.com is 40+ ms)
Will have to see if I get anywhere this week.
The monitoring device has been collected, I think it’s possible it’ll show the issue, there was definitely some evidence yesterday 8-10pm although I think that could be something else since the same isn’t shown for the other days.
https://i.imgur.com/mIltv9E.png (zoomed https://i.imgur.com/1rX2S5h.png)
The engineer mentioned that the issue had been marked as resolved which was news to me, I’ll have to see how it is this week maybe yesterday was an unrelated blip.
I’ll probably give them until Thursday to analyse the data and contact me, else I’ll chase them up.
If it does appear to be resolved after testing it a few nights this week there are a few things I would like to know:
1) What was the issue and is the current upcoming maintenance related?
2) How widespread was the issue? (e.g. all customers, certain exchanges, specific exchange equipment, only some packages, etc.)
3) How long has it been an issue? (I don’t think they’ll tell me in case it prompts compensation claims, but can they see that a change was put in place I think sometime around September-November that caused the problem)
Nice one! Good to see it did capture some issues for them to look at.
I’ve heard nothing at all, maybe they no longer need me to use the pi if they’ve classes it as resolved?
Appreciate you pushing and keeping us all updated!
It has taken a month but I’m rather confident that the current essential maintenance has resolved the problems for me.
My monitoring shows that at 2am the PPPoE session briefly dropped and this morning I could see that something had changed because the remote IP of the PPP link (e.g. the gateway) was now a public IP, and the first few hops after it on a traceroute are also different (10.78.*.* instead of 10.75.*.*).
So, I was hopeful that whatever changes were put in place were to resolve the issue and tonight for the first time in 4 months I’ve streamed ~2 hours of BT Sport 4K (~30GB) at peak time without any issue, no buffering, and the app device checker passes instantly.
All speed tests are reporting full speed.
I’ve no idea if my pestering or evidence-based issue reporting made any difference, I expect collectively they just received enough complaints to investigate the issue then once they identified the problem it took time to plan how to resolve it.
It does suggest to me that as I hypothesised there was a bottleneck at peak times, which required a redesign of the network to resolve.
What irritates me is the problem seemed to go on for 4 months, do they not have monitoring that would have shown the issue? (e.g., congested links or overloaded network devices)
Once I’m sure it’s fixed, I might see if I can get any details out of them for my own curiosity 🙂
For reference I’m on the exchange for the HU10 area, they posted about 7 nights of maintenance and as far as I know there are 14 exchanges so they could be changing 2 per night (I’m assuming they terminate the fibre at the same old exchanges).
Anyway, keep an eye out for any signs the maintenance has happened to your connections and let me know if you see an improvement.
Cheers Miken. I’m HU7. I’ll keep an eye out!
So all maintenance should have been done according to their site.
I’ll reboot the router today and see how things go.
Anyone had anymore issues lately? Last few weeks I’ve seen random dropouts, poor Ping, packet loss and slower than usual download speeds (Friday night was a joke)
Contacted them and they told me to change my ethernet cable. Safe to say I ended the chat after that!
KCOM is pushing out to new areas, but not backfilling in bypassed areas. I left KCOM to get 75 Mbps with BT (cabinet). Returned to BT for the same service. This has now been downgraded to 50 Mbps (email stated that they were improving the provision somehow!). Constant harping on about their brilliant “full fibre” service, but apparently not available here (although it was shown as being in place to several close properties on an availability map .. this has quietly disappeared after my enquiries). I will be returning to BT land shortly, where I at least feel that I am not lied to about the provision available.
*Returned to BT –> returned to KCOM