0.5 Mb Current Line Speed - Two PCs?

[#1]

Maybe it's coincidence and totally unrelated? - as I've just been 'jiggling' between two PCs. But I seem to have been landed with a very similar problem to that reported by Fozastuta in this recent topic

I've been on Plusnet Value package since 2009-06-22 when my "Estimated line speed" war recorded as 6Mb. I may be mistaken as I haven't had any need to check the Plusnet High-speed Broadband link very often, but I don't ever recall seeing additional info pertaining to "Current line speed". I've always had very reliable connections with consistently fast speeds. But yesterday my speed suddenly dropped drastically, and I observe (like Fozastuta) that my Current line speed is now recorded at just 0.5 Mb.

To paint in the background. On Tuesday last (22nd Feb) I purchased a much needed (failsafe) second computer. As I don't use wireless, I spent much of that day interchanging my ethernet cable connection between the two computers, and also had occasion to update the Realtec Network drivers on the new PC. I wouldn't have thought that would interfere with recorded line speeds, but yesterday was when I first became aware of a very severe slow down, and *I'm thinking* it may not be coincidence that I spent most of yesterday connecting with the new PC system again?

Whilst I don't pretend to understand all the techie stuff related to connection speed, I'm happy to conduct tests and supply further information if requested. Plusnet Tools informs me that I'm currently connected to the gateway known as ptn-ag01, and VP capacity at my exchange is showing as Green with no ongoing Service Outages. Maybe it's related to my problem, or just a busy time right now, but unfortunately I'm unable to conduct a BT Speedtest at the moment: The Performance Tester is currently unable to run a speed test for your broadband connection. Please try again shortly, however if this problem persists, raise the issue with your service provider.

Thomson TG585 v7 Current Link Information:
Uptime: 1 day, 7:40:42 
DSL Type: G.992.1 annex A
Bandwidth (Up/Down) [kbps/kbps]: 448 / 8,128 
Data Transferred (Sent/Received) [kB/kB]: 0.00 / 0.00
Output Power (Up/Down) [dBm]: 12.0 / 19.5 
Line Attenuation (Up/Down) [dB]: 8.0 / 16.0
SN Margin (Up/Down) [dB]: 24.0 / 13.5 
Vendor ID (Local/Remote): TMMB / TSTC
Loss of Framing (Local/Remote): 75 / 0
Loss of Signal (Local/Remote): 27 / 0 
Loss of Power (Local/Remote): 0 / 0 
Loss of Link (Remote): 0 
Error Seconds (Local/Remote): 30 / 0
FEC Errors (Up/Down): 0 / 588 
CRC Errors (Up/Down): 0 / 0 
HEC Errors (Up/Down): 0 / 0

My router is turned on 24/7, and I've always powered down both of my PCs before interchanging the ethernet cable. However, even if I/we can resolve this ongoing speed issue, I'd appreciate some refutation or opinion as to whether there's any possibility that whatever is the root cause of my 0.5 Mb current line speed, is not likely to re-surface whenever I have need to interchange the cable again?

Thanks for any help and elucidation.

[#2]

These are my immediate thoughts. Apologies if they seem unduly terse.
1. The "Estimated line speed" is demonstrably inaccurate. Best to ignore it.
2. The "Current line speed" is the line's IP profile, which is a somewhat artificial cap on the line's actual speed.
3. The BT speed tester failure is perhaps a separate problem; join the club!
4. Your downstream bandwidth (sync speed) of 8128 kbps implies an IP profile of 7150 kbps (source). If your actual IP profile is just 0.5 mbps (500 kbps) then your question is justified. Whether Plusnet can explain such a speed degradation is another matter. However they will suggest that you leave your router running for a while (hours, days) to see if the IP profile recovers.

[#3]

Have you only got a single port router? If not why can't you just keep both PCs connected and stop all the messing about?

[#4]

Thanks for your comments, linux
jelv - I've only got one ethernet cable, so I've been forced to interchange. I expect I'll probably purchase a second cable soon, but hey I only got the second PC a few days ago. My question remains valid however - I'm surprised if interchanging a single lead between two PCs has potential to influence reported line speed. Are you implying that it does?

[#5]

I wondering if the wire from the router to the filter is getting disturbed when you swap over (pure speculation).
Keep trying the BT speed test - getting the results from that would be really useful.

[#6]

Quote from: jelv

I wondering if the wire from the router to the filter is getting disturbed when you swap over (pure speculation).

No - definitely not

Quote from: jelv

Keep trying the BT speed test - getting the results from that would be really useful.

Will do.... Incidentally,  I don't really understand how the BT speed testing works. Given that it's currently Sunday afternoon - a very busy time on the net generally, as well as introducing some restrictions with Plusnet Value package speeds - will that influence or impair the results I can expect to receive?

[#7]

Plusnet prioritise the BT speedtest, but in any case it's the IP Profile speed that it reports that we really want to see.

[#8]

Hmmmm. I'm still left a bit confused to be honest jelv. To rephrase my earlier question in this context. Is it a possibility that if I was to conduct a BT speed test speed test right now during this busy weekend period then the IP Profile for my line that the test reports might be 'unduly' inferior to more typical results obtainable from tests conducted at less busy times?
I hope you can understand my concerns. Reflecting further upon linux's comments about somewhat artificial caps on the line's actual speed leaves me a bit concerned as to whether Plusnet's eventual resolution of my problem could in any way be unfairly influenced if using as it's basis only the results from a single BT speed test, which I may perform during this busy weekend afternoon period? I can only repeat that prior to this experience, I've always had consistently fast and reliable connections. I certainly wouldn't want Plusnet to find reason to begin imposing artificial caps on my future speeds because of one (potentially untypical and inferior?) test result.
Apologies if my worries are unfounded.

[#9]

Your IP Profile does not depend on the time of day, or how busy the network is at the time, nor is it an artificial cap (though many LLU providers handle the process better than BT). The IP profile is set (by BT Wholesale) based on sync speed events and to get a profile of 0.5Mbps (500kbps) infers your line must have synced with the exchange within the speed range 576-832kbps. Now that your sync speed has increased again to 8128kbps your IP Profile should go up to 7150kbps.
This increase should occur within 1-2 days (it should be quicker than that but the BT side seems to be taking longer these days), and it might well have happened already on the BT side (which is what a BT Performance Test gives). Once BT's IP Profile changes the new value is notified to Plusnet and should be reflected in Current line speed within 12 hours.
Unfortunately this automated process sometimes breaks, but without the BT Performance Tester result users cannot tell. However Plusnet staff can do a line test to check. Hopefully the change will have worked its way through the system by Monday morning when the Digital Care Team are around again.

[#10]

Quote from: spraxyt

Your IP Profile does not depend on the time of day, or how busy the network is at the time, nor is it an artificial cap ...

I stated that the IP profile is an artificial cap because it seemingly has nothing to do with the line rate, line stability, or anything else. It is purely a software cap on the line's throughput. BT's own explanation for the IP profile is incoherent.

[#11]

I'm very grateful for the helpful info spraxyt, as well as the additional food for thought from linux. I'll wait patiently to see what happens to my profile. But I have to say that even if I do see an improvement soon, I'll probably still remain a bit concerned about whether the root cause of the reduction in my IP Profile could indeed be in any way related to simply plugging my ethernet cable into my alternative PC system (and back again) - and whether slow sync speeds might be triggered again when I next do so? Even after I get a second ethernet cable, it's still going to be necessary for me to power down my main PC as I'm using the same monitor, keyboard, mouse, speakers for both - so there really won't be all that much difference?

I accept it seems unlikely that the problem is rooted here, but it does seem more than coincidence that after nearly a couple of years without any issues whatsoever such a drastic reduction in sync speeds has 'apparently' arisen precisely when I've started using two PCs to connect with (prior to this occurrence I think I've consistently had 448 / 8,128 and unless Plusnet disconnects me typically my uptime extends to many continuous days - sometimes 100+ IIRC)

(I've now tried to use the BT Speedtest several more times, but it persists in giving me the same test error. I think I'm a fully-fledged member of the club now!)

[#12]

Quote from: caulbox

it's still going to be necessary for me to power down my main PC as I'm using the same monitor, keyboard, mouse, speakers for both - so there really won't be all that much difference?

What you need is a KVM switch like this. [1]
I have 4 different computers sharing the same keyboard, monitor and mouse through a similar device.
[1] Not recommending this particular model, it's just an example...

[#13]

Quote from: linux

I stated that the IP profile is an artificial cap because it seemingly has nothing to do with the line rate, line stability, or anything else. It is purely a software cap on the line's throughput. BT's own explanation for the IP profile is incoherent.

The IP Profile is very much dependent on sync speed, it will be 7150kbps for a sync speed of 8128kbps for example, but that's all the IP Profile depends on. The DLM changes target noise margin or adjusts interleaving depth if, in its wisdom, it perceives the line as unstable (frequent disconnections) or the error rates on the line are significant. Increased target noise margin reduces sync speed, and that change can reduce the IP Profile as a secondary effect.
The BRAS profile is larger than the IP Profile, it is actually 7320kbps for an IP Profile of 7150kbps. So what BT say in the quote is correct.

[#14]

Quote from: Jaggies

Quote from: caulbox it's still going to be necessary for me to power down my main PC as I'm using the same monitor, keyboard, mouse, speakers for both - so there really won't be all that much difference? What you need is a KVM switch like this. [1] I have 4 different computers sharing the same keyboard, monitor and mouse through a similar device. [1] Not recommending this particular model, it's just an example...

Hey that does look interesting. Seems a very reasonable price too. Thanks!

[#15]

Quote from: spraxyt

The IP Profile is very much dependent on sync speed, ...

Perhaps we're confusing actual and theoretical IP profiles. The actual IP profile is obviously limited by the sync speed, but it is not dependent on the sync speed. In other words the two do not move in lockstep. If they did then the OP would not be experiencing throughput problems; his actual IP profile of 500 kbps is nowhere near the theoretical IP profile of 7150 kbps allowed by the 8128 kbps sync speed. Hence the actual IP profile is an artificial cap; the bits travelling down the line are still travelling at 8128 kbps, but the software is capping the IP payload to just 500 kbps instead of allowing the theoretical 7150 kbps.

Quote from: spraxyt

The BRAS profile is larger than the IP Profile, it is actually 7320kbps for an IP Profile of 7150kbps. So what BT say in the quote is correct.

The corresponding BRAS profile and (theoretical) IP profile figures differ slightly. But I don't see how that explains the BT quote, which talks about discarding and resending data.