drj - speed problem split off

[#1]

Not sure how relevant connected and unconnected errors are - but pro-rata it's not significantly different to a typical reading from previously

[#2]

Quote from: drj

I've have the router plugged in to the BT master socket for several weeks

Do you mean the test socket behind the faceplate?

Quote from: drj

I've now relocated the router on the professionally hard-wired extension

Don't wish to be a pain but that still could cause issues. We can't raise faults while a connection is in an extension socket- which isn't to say it's not ok, just that it has to be ruled out.
If you were in the test socket I would reset the SNR, but as you're not it could go back to 15dB and we wouldn't be able to do anything about it for 10 days.
Jojo

[#3]

Quote from: Joanne

Quote from: drj I've have the router plugged in to the BT master socket for several weeks Do you mean the test socket behind the faceplate?

As I said before:

Quote from: drj

I don't have a test socket. It's currently plugged into the master socket on an extension (since the socket's in the kitchen) ...

The wire comes from the telegraph pole in the road across to a box on the top of the wall outside. From there a wire runs down the outside of the house, through the conservatory and into the kitchen. At that point there is a master socket. It is a conventional master socket - there is no test socket inside it. Would you like a photo?.
From the master socket there is what appears to be a professionally-wired 6-wire extension up to a wall-fixed socket in the sitting room upstairs.
I had the router plugged into the kitchen socket but that means the router has to be on the floor by the kitchen door - that's totally impractical. The speed when it was plugged in there was no more and potentially less than it had been when attached to the end of a loose extension lead I'd rigged up into the dining room (which still had to go through the kitchen door meaning I couldn't close it).
I cannot have my router in the kitchen because my main computer is network cabled into it. Also there is no mains power in the vicinity of the mater socket so it would mean an extension lead across the kitchen!
I want the router upstairs to give good coverage - the house is split level with a substantial storey height retaining wall potentially blocking the signal if the router is downstairs. I've previously been running the router on top of a high shelf unit in the downstairs dining room next to my computer to raise it as high as possible.
With respect, although I can understand where your coming from, what is the point in testing and resetting the whole system when I don't want the router plugged directly into the master socket?
Are you really saying the only solution is to pay BT lots of money to come to my house and re-site the master socket (which I presume with then be a test socket version) upstairs?

[#4]

Why in the master socket you ask. Well if you can prove that the fault is still evident while connected to the BT master socket, then BT are obliged to fix the issue (if they can) free of charge to you. This is because BT are responsible for the wiring from outside up to the master/test socket.

[#5]

HI drj,
As RPMozley says, it's only so we can be sure that the fault is not internal. Once we do that we can raise it to our suppliers in the knowledge that it must be on their network.
Jojo

[#6]

But we haven't established there is a fault yet have we. I'm merely trying to optimise the speed.
Would you please explain in simple language what you think the problem is?
Please also note that I didn't reply yesterday since the notification has only just been delivered into my mailbox overnight. Refer screenshot. You can clearly see I downloaded yesterday's emails up to about 9pm (with 10 new ones having come in since then) but the two notifications on this topic was obviously severely delayed. This is where SM comes in useful since that shows when emails arrived not when they were sent - after 4am this morning in this case.

[#7]

All right - I'll put it all back to the master socket - but why's it SLOWER when plugged in there than when having an extension?

[#8]

Off topic:

Quote from: drj

This is where SM comes in useful since that shows when emails arrived not when they were sent - after 4am this morning in this case.

In Roundcube when you have a message displayed (either by opening the message or in the preview panel) you can click the down arrow below the headers to display the full headers. That shows the delivery date.

[#9]

Hi drj,
I could reset the SNR which would improve the speed. However, you would have to leave it in the master socket. If I reset it and it's not in the master socket or you move it out of the master socket afterwards the SNR may go back up. Once I initiate the reset it will start a 10 day training period, so you wouldn't be able to raise any performance issues during that time. Let me know if you want me to go ahead.
Jojo

[#10]

Right - I've now worked out what's going to be the optimum temporary situation - I've plugged the router directly into the Master socket (via the same filter as I was using upstairs) and sited it in place of a picture in the dining room on the other side of the wall from the master socket (with the lead going through the door opening) powered from a mains socket underneath. At least hung on the wall I'm not going to step on it where it was previously nearly in the door. My main computer is plugged into the router via a long extension. The router is reporting lower up and down speeds than it was before - which parallels previous experience - see screenshot 1.
I've tried doing speed tests at speedtest.net, but unlike previously when I was generally getting a flat response, this time FF was consistently   showing a high peak about 3/4 through the download test which otherwise was hovering around 2.0 as usual - I tried several times with the same result - see screenshot 2. Hence I closed FF down and tried Chrome - that was extremely peaky - see screenshot 3. Then I tried Internet Explorer and that was merely slow (consistently less than 2). Then finally FF again and again it was peaky - see screenshot 4. So the theoretically achieved results are to be viewed dubiously.
However does that mean that with a stable connection I should be able to get around 2.8 rather than the 2.0 I usually get? That might make streaming video watchable which it often isn't presently.
In comparison I was sometimes getting in the low 3's when on Orange - but that could crash daily or so.
Over to you please...

[#11]

You've just resynced the router during the evening - it's perfectly normal to see lower speeds when you resync around this time. If you resync tomorrow between 10am and 3pm you should see a higher sync.

[#12]

Thanks - I presume "resync" means turn off and back on? I'm at work tomorrow...

[#13]

Right, I'll reset the SNR and we'll see how it goes

[#14]

@drj:
JoJo's reset should have resulted in a resync - you should be able to work out when from the up time in the Wan Status. If that was during daylight don't do another one - just leave it well alone and on 24x7 to see where it settles.

[#15]

Thank you - the figures presently look much different (see first attachment).
I shall see how things go on - however repeated speedtest.net results were very peaky (see second + third attachments) and even the peak was lower than yesterday:


Let's see how things are over the next few days (note I may have to go away over the weekend)