When congestion is given as a reason for speed drops, how much is acceptable?

[#1]

@Gandalf I noted an improvement over the weekend. There was still a drop at peak times, but the base line has moved up from <3Mbps to >10Mbps.

I'm keeping my logging software on the case for the next couple of weekdays - which previously showed the sharpest declines - just to make sure.

[#2]

I believe Gandalf is on leave this week so you are unlikely to get a response. It still sounds like you are on a congested SVLAN, you should not be getting a significant speed drop at peak times. I expect all the links are congested. Unfortunately this may take a lot longer to fix if additional equipment is needed at the exchange. Just keep on monitoring so you have the evidence.

[#3]

And I'm back.

I'm keeping my logging software on the case for the next couple of weekdays - which previously showed the sharpest declines - just to make sure.

Keep us posted on this.

[#4]

Hi @Gandalf - I stopped logging as 'user complaints' in the household went silent. Just for giggles I'll fire it up for one more 24 hour period and get back to you.

[#5]

I've reviewed weekday performance again. Before we sign off on my speed issues would anyone care to review this chart and offer an opinion please?

[#6]

I stopped logging as 'user complaints' in the household went silent.

That's always a good sign that things are working as they should. 

I've reviewed weekday performance again. Before we sign off on my speed issues would anyone care to review this chart and offer an opinion please?

It looks like you're getting maximum speeds during the day and throughout the night, but dipping slightly in the evenings.

This may be caused by the connection being used.

Do people in your household generally use the internet more in the evening?

[#7]

That's the thing @Gandalf - on the first day shown, yes, my son was probably giving his PS4 a hammering. However, on the second day we were all out for the evening. I have no 'automated' devices that access the internet to that extent.

I'm just curious at this stage - dropping from 39Mbps to around 25Mbps still has to be localised congestation surely? Assuming yes, am I going to be seeing significant slowdowns again as the fresher VLAN becomes populated, and is there a process whereby Plusnet as the customer of BT Wholesale formally requests better investment at the exchange?

[#8]

Sorry about not getting back to you in a timely manner.

I'm just curious at this stage - dropping from 39Mbps to around 25Mbps still has to be localised congestation surely?

It's possible.

am I going to be seeing significant slowdowns again as the fresher VLAN becomes populated

If the fresher VLAN gets to a stage where the max throughput exceeds the threshold where it becomes congested and things start slowing down, you might feel slowdowns. The VLAN you're on now is still showing as green.

With that said, the previous VLAN was also showing as green so if you're experiencing these problems still please can you let me know and I'll speak to our suppliers quality assurance team again?

is there a process whereby Plusnet as the customer of BT Wholesale formally requests better investment at the exchange?

Not that I'm aware of.

[#9]

Hello @Gandalf - thanks for the reply.

I have more information for you. My throughput has indeed fallen off again - admittedly not as much as last time, but the trend is clear. Graphic attached. Daily slowdown occurs around what I guess is peak family time of 7pm - with some serious drops to <7.5Mbps.

So, despite the VLAN still 'showing green' as last time, I'd bet this is going to get worse again.

Can you escalate this to your supplier's quality assurance team please?

(I hope no-one asks for another round of OpenReach basic testing at this end - nothing has changed here; and this is supported by the fact that maximum throughput is achievable during some parts of the day)

Thanks.

[#10]

Seeing how badly Plusnet have handled this case totally validates one of my mail reasons for moving to AAISP. They would have recognised (as did many of the experienced users on here) that this was a cabinet/exchange capacity issue straight away. In fact AAISP's line monitoring would almost certainly picked this up before you even reported it. Not being part of the BT group they would have been far more firm with BTw and you wouldn't have had the pointless engineer visits or router replacements.

[#11]

Thanks @jelv - I can't disagree with any of the points you raise given my experience with this and the Safeguard debacle. I am in the position where I recognise that 'quality probably costs' and, at less than £30 per month including line rental, perhaps I'm expecting too much of PN. But I can't justifiy extra expense. Call me old-fashioned though; I do believe I should get what I pay for.

I've been with PN since the Free-Online days - circa 2000 I'd say.

[#12]

Sorry to hear you're still having issues @DocDelete

Can you run a BT speedtest at http://www.speedtest.btwholesale.com using a wired connection, completing the further diagnostics at the end? This will store the result on the supplier system and will help us in reporting a fault to our suppliers.

[#13]

@DocDelete Make sure you run the test Gandalf has requested when the line is running really slow - do not run it when it is OK because if you do BTw will say NFF!

[#14]

@Gandalf @jelv Well, yes of course I can run the test - and clearly I will try to do so when the VLAN is congested. I'll give it a whirl later.

But what happens if that /single/ test falls between troughs and I get a good result though? Am I expected to keep repeating until I massage the result to where I want it?

I'm sorry - but experience is making me question all of the hoops to jump through.

The further tests demand a level of technical competency that I have, but I can imagine many users won't possess. Surely there's a way for the line throughput to be interrogated from the other end too? Having to build a case seems a bit too close to doing someone else's work for them.

[#15]

---

@DocDelete wrote:

Thanks @jelv - I can't disagree with any of the points you raise given my experience with this and the Safeguard debacle. I am in the position where I recognise that 'quality probably costs' and, at less than £30 per month including line rental, perhaps I'm expecting too much of PN. But I can't justifiy extra expense. Call me old-fashioned though; I do believe I should get what I pay for.

I've been with PN since the Free-Online days - circa 2000 I'd say.

---

Yes, AAISP would cost you around an extra £200 a year depending on how much you download and whether you use the phone. But how much is it worth not having to wait in for four pointless engineer visits, all your time testing, retesting, waiting on the phone for someone to answer and posting on here? On top of that how much is it worth to get rid of the earache you get from the rest of the household when the internet is slow?