30-06-2018 10:09 AM - edited 30-06-2018 12:31 PM
My wifi hub has now been installed and running for a few days with DLM making no changes, so some sort of interesting stats from the gui that an OCE may want to feed back to the development team.
Downstream noise margin is being reported as 0.00 dB.
Upstream attenuation is being reported as 0.00 dB.
Downstream attenuation is being reported as 25.90 (usually been in the 19-21 range with previous TalkTalk super routers).
Upstream max rate exactly 6000 Kbps, actual upstream rate exactly 6000 Kbps, upstream noise margin exactly 6.0 dB. Now these numbers could be correct, but they do look unusually rounded. With HG633 and HG635 my max. upstream was just short of 8 Mbps, so lost performance.
But the thing that concerns me is the performance of this "best ever" modem-router.
I have listed the upstream performanc loss already.
With an HG635 my max. downstream was around 49 Mbps to 50 Mbps. With an actual line rate of 40 Mbps (on the 40/10 package).
With an HG633 my max rate dropped to around 43 Mbps, with an actual of around 38 Mbps.
With this new wifi hub my max. rate has dropped to 35 Mbps with an actuall of 32 Mbps. This unexpectedly large drop in downstream performance could be linked to the unexpected rise in the downstream attenuation.
But having spoken to someone in the faults team they are happy with this performance level as it is still within my predicted range (but massively less than I used to get). It would appear that everytime TalkTalk introduce a "new, improved" router that my performance levels go south.
But the wifi is better, although I never had any wifi issues, and this new router didn't resolve the issues I was having with the HG633 with V2.00t firmware.
on 19-07-2018 08:40 AM
on 18-07-2018 04:49 PM
I think "probably" may be a bit of a stretch.
My router was crashing and resetting (according to the router logs) at more or less exactly the same time as my router was being remotely accesssed. I stop the TR069 access and there hasn't been any incidents since. A possible coincidence, but quite a long way from a probable coincidence.
A fact that may be worth passing on to whoever is doing the investigation rather than just "bearing in mind". Effects normally have a cause and disabling TR069 may just be the cause of the change in the reported error rates on my router/line.
It could all be indicative of a potential firmware issue.
on 18-07-2018 11:06 AM
on 18-07-2018 09:22 AM
Working on a possible theory that the misreported stats and my line crashes and retrains may be associated (with V2.00t firware changes) .........
To try and assist the investigation - I asked yesterday if there had been any issues indicated on my line in the previous three days, you said not. My router stats today are not indicating any further issues (which I guess you can confirm), so my line appears to have run clean now for around 4 days.
4 days ago I disabled TR069.
Might be just a coincidence.
on 18-07-2018 09:03 AM
on 18-07-2018 08:56 AM
The Royal Mail, more known for outdated business practises and underwhelming IT systems, can take my item on a Friday and over a weekend (non-business days) can have it physically at your address by first thing Monday morning, but here we are 2 days later and TalkTalk haven't manged to enter it onto your system.
The production team don't think a discrepancy between 2 major reporting metrics which should be identical is "not significant".
TalkTalk never fail to amaze.
on 18-07-2018 08:29 AM
The router isn't showing as returned yet, when it is we'll process the refund, can you bump the thread on Friday and we'll check again
Our product team have confirmed that they don't know why there is a difference between the sync speed reported by the router and the Openreach website. they don't really seem to think that it's significant
on 17-07-2018 04:46 PM
The no retrains since the 14th is quite interesting, I made a change to one of the routers parameters on the 14th, but still early days.
I will await the technical teams response with anticipation.
Seems like maths may not be my strongest suit either, but even by your lower suggested value I believe we are still within the realms of statistical significance. But if you consider that I was getting around 32 Mbps a few days ago, and 40 Mbps a few months ago it is still a significant performce drop.
Any progress on the promised refund ?
17-07-2018 02:51 PM - edited 17-07-2018 03:19 PM
"A difference of over 10% in your reported "BTOR" sync speed and the routers' reported sync speed is significant."
I make it more like 5% but I could be wrong, maths isn't my strong point 🙂
"Chris, I'm guessing that your background is customer services oriented rather than technical."
"Can I ask the question that I have already asked @OCE_Ady twice with no specific reply. From the stats you have access to can you see if there have been any issues, outages, resets, DLM interaction etc.. over the last 3 days ?"
The last profile change was on the 14th and there were 12 retrains on the 14th to, none since.
I'll ask our product team about the difference in reported sync speed, see if they can cast any light on it
on 17-07-2018 01:59 PM
My speedtest results are within acceptable bounds for a fibre sync speed of over 27 Mbps.
A difference of over 10% in your reported "BTOR" sync speed and the routers' reported sync speed is significant. Could it be a strange little bug in the new V2.00t firmware on the HG633 (I never had any issues before moving to this firmware level). It is the firmware that negotiates the line speed with BTOR's fibre systems, it is the firmware that reports what I can see - TalkTalk's firmware.
Chris, I'm guessing that your background is customer services oriented rather than technical. Most techies would at least be curious (rather than dismissive) about this issue. But another guess - TalkTalk don't pay you to be curious.
If we (the end users) can't believe the stats we are seeing from the router what are we suppossed to believe ? When an OCE asks for line stats for diagnostic purposes do they believe the numbers we report back or do they assume the router reported stats are wrong and should be ignored (as you appear to be dismissing the stats I am reporting) ?
Have you even considered asking TT's technical team for an opinion or to see if they have seen this before ? Not just does the line look OK, but the specifics of this particular problem. Inconsistances can be symptomatic of actual or potential problems. It doesn't need BTOR at this stage.
Something is not quite right, can you at least admit that ?
Can I ask the question that I have already asked @Ady-TalkTalk twice with no specific reply. From the stats you have access to can you see if there have been any issues, outages, resets, DLM interaction etc.. over the last 3 days ?
on 17-07-2018 01:20 PM
I've checked the sync speed directly on the Openreach website and it's showing 25.8Mbps. I agree that your wireless speed test results are high for a sync speed of 25.8Mbps, they are also high for a sync speed of 27.2Mbps.
I don't know why there is a slight difference between the sync speed reported by the router and the sync speed reported directly on the Openreach website or why the speed test results are so high but this isn't really something that we can report as a fault.
17-07-2018 11:42 AM - edited 17-07-2018 11:51 AM
Could I question the sync speed you are reporting (25.7 Mbps), I did ask @Chris-TalkTalk earlier in the thread about similar results but he didn't really answer.
You say my sync is currently 25.7 Mbps. I have just run three seperate speedest.
1. Ookla as provided by Microsoft shop - downstream 25.5 Mbps. This is 99% of your reported sync speed.
2. Thinkbroadband - 25.3 Mbps - 98% of your reported sync speed.
3. TalkTalk test via "My Broadband" (so Ookla again) - 25.4 Mbps - 99% of your reported sync speed.
So, 25.5, 25.3 and 25.4 - these seem to be very comparible results from different sources.
These speed tests were performed on a wifi connected Windows 10 PC - I would suggest that these are practically impossible figures based on your quoted sync speed, the lose between a router and any connected device is more than 1 or 2 percent. TalkTalk usually quote double digit "loses", 1 or 2 percent is deserving of examination.
My router is showing a line speed of 27.2 Mbps - this is consistant with the reported speedtests, and the ratio (low 90%'s) is very similar to the results I have had for the last 3 or 4 years as a fibre customer.
Is there a problem at your side causing low sync speeds to be reported ? Could this possible misreporting be a cause of DLM's lack of positive on intervention on my line.
Alternatively could you suggest a reason for these quite extrodinary results ?
But as to the question in my previous post - do your line stats show any outages, resets or issues, more specifically over the last three days ?
on 17-07-2018 11:14 AM
The line stats look the same as they did when you had 30mbps synch speed, but it's currently 25.7mbps. The refund isn't showing on the system yet. If the warehouse team are busy there may be a slight delay getting that on there. I'll see if I can chase it up for you.
17-07-2018 09:00 AM - edited 17-07-2018 09:03 AM
p.s. How are my connection stats looking - look ok here, but no signs of DLM making any attempt to recover my previous performance levels (so back at square 1 again ?).
on 16-07-2018 01:55 PM
Hi iefbr14, the router return has yet been processed and it's not showing on your account yet. I'd expect it to take a day or so to come through. Please post back to remind me to chase this for you tomorrow.
on 10-07-2018 01:03 PM