Register (or login) on our website and you will not see this ad.
|
|
I have contacted support about the speed issues but have been told it’s my WiFi. Again same unit as my previous ISP. I suspect it’s due to BT Openreach being my carrier instead of TalkTalk Business which had LLU’d the local exchange. I suspect that’s how I’ve been able to get better connectivity in the past.
You really need to test using a wired connection. Wi-Fi is fairly fluid, it can work great one day and not so the next depending on interference or if it has decided switch channels etc.
I was originally provisioned to the incorrect LNS by A&A they did fix it but I still get speed drops.
I had the same problem 2 years ago and provisioned on to the wrong LNS, I'd have thought they had got this fixed by now, obviously not.
You could try connecting to the test LNS and see if that gives you better speed, just prefix your username with test- and reconnect. I see better single thread speeds when on that one presumably because it isn't loaded up with many customers. If its better you can assume the slow downs are not due to BTWholesale and backhaul
Edited by E300 (Wed 17-Apr-24 13:09:39)
|
|
|
Just for clarity I also have experienced the LNS stalling issues I was told they are looking into it and they had added NVMe drives to some of the LNS systems to get things resolved.
I have contacted support about the speed issues but have been told it’s my WiFi. Again same unit as my previous ISP. I suspect it’s due to BT Openreach being my carrier instead of TalkTalk Business which had LLU’d the local exchange. I suspect that’s how I’ve been able to get better connectivity in the past. I was originally provisioned to the incorrect LNS by A&A they did fix it but I still get speed drops.
Like I said when it comes to contract renewal I’ll be looking elsewhere.
If it was load/congestion related your max would still be line rate as congestion wouldnt be around the clock.
Do you actually use the net over wifi or ethernet? Expecting gigabit throughput over wifi isnt realistic, and you need good equipment to even get half way there.
Your 300 does smack of provisioning problem as conveniently thats around the figure people report when OR get it wrong, and its also not far off what people reported when on the old AAISP LNS. But this is moot until you actually test over ethernet.
Edited by Chrysalis (Wed 17-Apr-24 13:57:58)
|
|
|
I have contacted support about the speed issues but have been told it’s my WiFi. Again same unit as my previous ISP. I suspect it’s due to BT Openreach being my carrier instead of TalkTalk Business which had LLU’d the local exchange. I suspect that’s how I’ve been able to get better connectivity in the past.
You really need to test using a wired connection. Wi-Fi is fairly fluid, it can work great one day and not so the next depending on interference or if it has decided switch channels etc.
I was originally provisioned to the incorrect LNS by A&A they did fix it but I still get speed drops.
I had the same problem 2 years ago and provisioned on to the wrong LNS, I'd have thought they had got this fixed by now, obviously not.
You could try connecting to the test LNS and see if that gives you better speed, just prefix your username with test- and reconnect. I see better single thread speeds when on that one presumably because it isn't loaded up with many customers. If its better you can assume the slow downs are not due to BTWholesale and backhaul
Whats your single threaded like on the production LNS? I am used to max single threaded around the clock, so would be disappointed if its like Zen.
|
|
Register (or login) on our website and you will not see this ad.
|
|
|
This is the current test, a bit better than it has been on the single thread test. I can replicate similar numbers by doing a single thread test at Speedtest.net.
IPv4, consistently around half the speed on a single thread.
https://www.thinkbroadband.com/speedtest/17133601033...
IPv6, which typically used to give lower figures due overheads is now better than IPv4, so something is a bit odd somewhere just recently
https://www.thinkbroadband.com/speedtest/17133599514...
The test LNS did a lot better on the single thread tests, 700+ from memory, but I can't disconnect and try it again at the minute.
It doesn't really matter now as I'm migrating away, the issues have gone on too long and the premium I'm paying isn't justified in the service received anymore. It is clear we are beta testing their new kit and paying them a premium to do it, I've stuck around and given them the benefit of the doubt but 5 or 6 months now and still not fixed.
Okay we are running back on stable firmware, but it still means another round of upgrades to come and potentially the same problem then rolling back again in the near future, then it all starts again a few weeks later. I think if they can't fix it after all this time, even when they have a stable firmware they should be able to compare and see what code changes have taken place, suggests this problem may not be fixable anytime soon and is quite a low level hardware issue.
|
|
|
Thank you, this combined that you see it performing better on the test LNS does indicate something not been right.
I am letting my order progress, but I wont lie my eyes have already wondered. Hence the post I made a few posts back.
I agree also on the hardware. Over the years I have done my own far share of diagnosing, kernel panics and the like, and it is good AAISP have been open, but when I read about CPU deadlocks, and it happening on all of the units, I cant look past either some kind of hardware issue (VRM or whatever) or a bios programming issue, the fact the factory firmware seems to be stable, but at the cost of not fully utilising the hardware gives hope a bios or p-state/c-state type adjustment could stabilise the entire kit and if I was AAISP that is what I would be pursuing, I have managed to make unstable kit stable via those kind of tweaks. But of course thats down to AAISP to figure out, I was surprised at the resistance to utilising cisco as a stop gap, and no compensation got mentioned, the path they ended up taking only served to hurt both brands I think. It is good at least they have now pulled back to the factory firmware though and left the test device as an optional one for customers to utilise. They at least got to that point in the end.
Curious, if you dont mind saying, who are you migrating to?
Edited by Chrysalis (Wed 17-Apr-24 15:19:05)
|
|
|
I was surprised at the resistance to utilising cisco as a stop gap, and no compensation got mentioned
Exactly, I've not heard of a plan B except they haven't got one, they are just blinkered into Firebrick and that's it, and I just got to the point of thinking it is too much money a month to be a beta tester. I agree they should have offered some rebate or reduced monthly premium whilst these issues continued to those affected.
After some research and some info from another user here I'm moving to Unchained. So far a couple of questions I've had have been answered in mere seconds it seemed, and its great to support a smaller business and get that more personal touch. It's cheaper than A&A as well, mind you most ISPs are!
A&A say they are transparent and happy to engage, but I think the last update on the problems only came about because of more noise in this thread, and despite saying they are happy to answer questions, questions asked of them here hasn't had any response I've noticed! They talk the talk...
I'm sure A&A will get it sorted out it in the end one way or another, and maybe I will return as a customer in future who knows. It was great for many months,couldn't fault it, but I can't justify to myself the premium for their product whilst these issues are ongoing with no end in sight.
Edited by E300 (Wed 17-Apr-24 16:02:05)
|
|
|
Thank you, will see how my experience goes when I rejoin, but I am preparing for the worst if it happens, and I appreciate you sharing.
|
|
|
Good point E300, aabloor the General Manager of A&A registered here on Fri 12-Apr-24 10:24:36 to post his message:
https://forums.thinkbroadband.com/aaisp/t/4755788-re...
he finishes that message with:
"Thanks for taking the time to read this, and we are happy to answer any questions, of course."
But no response has been made to ANY of the questions that have followed... and he still has:
Total Posts 1
|
|
|
Whilst there's technical reasons to keep PPP in the BTW and TTB networks that would be harder if it was a pure layer 2 delivery (at least in my opinion), they both do offer VLAN based handover on their EoFTTx products - which attract a fair premium, and they probably want to keep EoFTTx as their premium offering to maintain that extra profit margin, which would incentivise keeping PPP in place for their less premium offerings.
FTTx is already handed over on the L2S in the exchange as a VLAN anyway.
I don't understand: why would it make a difference how premium something is whether it's delivered via PPP or Ethernet? Surely it's the SLA and QoS behind the service that makes it premium, not whether it's delivered as Ethernet or PPP?
Ethernet over FTTC is FTTC with a service level agreement - https://daisycomms.co.uk/resource/what-is-ethernet-o... - but is presented to the customer as Ethernet. It could happily be a PPP service with the NTE stripping the PPP to give an Ethernet presentation. The presence of PPP costs a little throughput but beyond that it makes no difference to the service's reliability or performance.
On it being harder not really, just different. CityFibre provide Ethernet handoff to everyone that wants it with PPP as an option. Broadband Network Gateways built to handle Ethernet presentation and handoff may eliminate the need for PPP, allowing providers to use EVPN technologies. Still encapsulating but a whole lot more you can do with the traffic than PPP over L2TP tunnels.
Edited by XGS_Is_On (Wed 17-Apr-24 20:34:22)
|
|
|
I was surprised at the resistance to utilising cisco as a stop gap, and no compensation got mentioned
Exactly, I've not heard of a plan B except they haven't got one, they are just blinkered into Firebrick and that's it, and I just got to the point of thinking it is too much money a month to be a beta tester. I agree they should have offered some rebate or reduced monthly premium whilst these issues continued to those affected.
Mentioned this in my earlier post to Alex: the ISP arm can't drop the Firebrick hardware as they're the largest customer by far and if the ISP that's part of the same group as the hardware company are using someone else it's not a great sales story.
With that in mind no surprise at all that AAISP have no intention of using anything other than Firebrick. The hardware was built with their requirements in mind, customised to them and relies on them for sales. These folks work out of the same offices and are mates and colleagues so why would A&A do anything other than keep the faith unless the issue were catastrophic?
I'm not being mean with that, I'm just, well, surprised at Chrys being surprised. I would've thought this would be obvious.
|
|
|
Pages in this thread: 
Print Thread
