Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[boost]

I have to say I'm beginning to regret joining Plusnet.

Web pages stalling etc.

I've seen this behaviour too. Their DNS is crap, presumably (my connection is otherwise flawless).
I've avoided changing to Google DNS to see if the issue persists but it might be worth a try for you?

[HighBeta]

Plusnet are going down the same route as BT with dns.

If you're still experiencing problems with blocked content, this may be due to you using a different DNS to Plusnet.

http://www.plus.net/support/safeguard.shtml

Would be wise to change settings while this is being implemented. As if its anything like the Mx960's "  Junos OS crack team " its going to be fun and games all round. 

[Chrysalis]

are you making assumptions? I dont see anything about enforced dns rerouting.

Also that safeguard is disabled on my account.

[tommy45]

Well the peak time congestion is still ongoing for some of their customer base, and they are still in denial that a problem  exists  they have started blaming customers connections suggesting that they must have a line fault, basically so far their support has been as much use as udders on a bull!! imo
I Still intend on migrating but not until i have explored the formal complaint route  all the way to ADR  if need be,

[boost]

I switched to Google DNS. The stalling web page issue has disappeared for me.

[jelv]

See http://forum.kitz.co.uk/index.php?topic=15087.msg281445#msg281445

This is not just a Fibre or even a 21CN issue, I'm seeing classic congestion symptoms (single stream must slower than x6) on 20CN Max Premium.

There's a significant fact here: Max Premium has a higher priority on the BT network. So that leaves one conclusion - it's Plusnet.

Oh, if anyone suggests it's a line fault causing a high error rate, my current stats:

Uptime:   5 days, 13:10:12
DSL Type:   ITU-T G.992.1
Bandwidth (Up/Down) [kbps/kbps]:   832 / 8,128
Data Transferred (Sent/Received) [MB/GB]:   752.71 / 4.34
Output Power (Up/Down) [dBm]:   12.3 / 19.9
Line Attenuation (Up/Down) [dB]:   11.0 / 20.0
SN Margin (Up/Down) [dB]:   15.0 / 6.5
System Vendor ID (Local/Remote):   TMMB / ----
Chipset Vendor ID (Local/Remote):   BDCM / TSTC
Loss of Framing (Local/Remote):   0 / -
Loss of Signal (Local/Remote):   0 / -
Loss of Power (Local/Remote):   0 / -
Loss of Link (Remote):   -
Error Seconds (Local/Remote):   26 / -
FEC Errors (Up/Down):   0 / 0
CRC Errors (Up/Down):   16 / 27
HEC Errors (Up/Down):   - / 24

[jelv]

Is it a co-incidence that the peak time issues which suggest under investment in capacity started appearing around the same time as Plusnet started going crazy with cash back offers?

[tommy45]

You must surely mean the race to the bottom? yes i have noticed a concerning trend in a downward direction, and not only jitter at peak times but packet loss during those times is becoming more common place, and i get the feeling they don't care, staff will be told to keep shtum and do what they have so far blame everyone and everything apart from their selves

[boost]

See http://forum.kitz.co.uk/index.php?topic=15087.msg281445#msg281445
I'm seeing classic congestion symptoms (single stream must slower than x6) on 20CN Max Premium.

Is this really classic congestion? If there's only 10% of a pipe left and you launch 6 * 10% threads, do you get 60%?

At best (worst?) this is some sort of congestion avoidance/management. Shaping and/or weighted tail drop perhaps? I think it's important to draw these distinctions if we are to truly understand wtf is going on here.

Or we can peddle myths? That could be fun too

So... let's assume this is real. At what point, EXACTLY, in the network might it be happening? What device?

[tommy45]

I think it's where BT's network meets the internal network of plusnets Handover point/s As if by hopping gateways  throughput is restored  sometime even if returning to the same gateway that you first saw congestion low throughput on,

I came across one endpoint whilst hopping gateways,offering a  sub 10mbps down and less  than 1 mbps upstream it also had a base latency of over 100ms, I'm on FTTC 80/20 With ip profile of 77.44 if it isn't congestion then something is seriously wrong with their traffic management , maybe that it's unable to function properly once customer usage reaches a certain level ? The BTW PT Tap 3 test will not resolve to the speed-test page  once i change log in credentials as per on-screen instructions , that's a convenience in plusnets favour too

But whatever the cause  plusnet should of fixed the problem long before now,instead of denying that it's their networking  by saying we cant find anything wrong, lame to say the least
 

[jelv]

@boost

When all is fine a single thread will run at near line speed and 6 threads will run at ¹/₆ line speed.

When there's congestion all threads for all users will run at say 50% speed. So a single thread speed test will be significantly impacted and a x6 speed test will still get near line speed. The other classic sign is that speeds during a test are up and down like a yo-yo.

So if it's not congestion somewhere what is it?

Outside peak times I can get a rock steady download (both single and x6) at around 94% of my IP profile. I've been told the status of my exchange/VP is good. Max Premium has higher priority on the BT network. The finger is pointing in one direction.

Also could you explain how on a bad evening some users find a significant improvement by gateway hopping?

[boost]

@boost

When all is fine a single thread will run at near line speed and 6 threads will run at ¹/₆ line speed.

When there's congestion all threads for all users will run at say 50% speed. So a single thread speed test will be significantly impacted and a x6 speed test will still get near line speed. The other classic sign is that speeds during a test are up and down like a yo-yo.

So if it's not congestion somewhere what is it?

Outside peak times I can get a rock steady download (both single and x6) at around 94% of my IP profile. I've been told the status of my exchange/VP is good. Max Premium has higher priority on the BT network. The finger is pointing in one direction.

Also could you explain how on a bad evening some users find a significant improvement by gateway hopping?

I think you're definitely onto something but I don't understand what exactly. Sorry if it came across as 'shut up foo, there aint no congestion!'

I feel pretty ignorant about the whole thing. I've been on the main site, looking at the diagrams and trying to piece it together but I'm a long way off

I've been told the status of my exchange/VP is good

What does this mean exactly?
What exactly is a VP?
How is it provisioned?
How many are there, typically?
How many at your exchange?
Who's definition of 'good' are you reporting? What
exactly is their definition of good?

These are the kinda questions that fill my mind when I read stuff on here.
IT has some brilliant people working in it but to say this is the norm would be extremely creative

Max Premium has higher priority on the BT network.

Says who?
Why?
How does this higher priority manifest itself on your traffic?
Assuming traffic, what happened when IP Stream became WBC?
Are you on a native IP Stream port? (I can see that you are)

[jelv]

Ages ago (pre ADSL Max) BT did residential products with 50:1 contention and business products with 20:1 contention.

When ADSL Max came out the equivalent products were Max and Max Premium.

Plusnet say it provides a prioritised service: http://www.plus.net/support/broadband/products/prioritised-service-faq.shtml (this page is for the prioritised service and says that on 20CN it also gives the faster upload).

See also http://en.wikipedia.org/wiki/ADSL_Max#Max_.27Premium.27

WBC isn't available on 20CN only exchanges!

[kitz]

I find it really weird.   For weeks I suffered really badly.  Speeds were down to circa 20Mbps at peak time.  I did all sorts of tests which were completely overlooked by Plusnet.   Then one night when things were bad I did a gateway hop and immediately went from 20Mb to 77 Mb.   Ive been perfectly fine since Ive been where I am now.... but Ive been on this pipe for several weeks and not disconnected.   I dunno how things will be if I have to resync or get a new PPP session.


 

[roseway]

Today I've had a similar issue, but with the upstream mainly. After a forced resync (for other reasons) web page loading became very sluggish so I did some speed tests. The BTW test didn't record an upstream speed at all, and others recorded upstream speeds of less than 0.5 Mbps. The upstream sync speed is 20 Mbps.

I dropped the PPP session and let it reconnect, and immediately everything was working normally again, with up and down speeds close to the sync speeds.