Any capacity issues?

[#1]

mine has been ok but I suspect its simply because i am on a good endpoint, if I hop around I will probably find bad endpoints.

[#2]

A good endpoint lol I'm beginning to think that that's near impossible last night i hopped from PCL BNG 02 onto PTN-AG01 As single threaded throughput was poor and there was also around 4% packet loss to all destinations at times this packet loss started from within plus nets remit from one of their visible links, I hopped gateways and the packetloss disappeared but there was some jitter, and throughput although improved wasn't 100% as it is outside peak times
Tonight once again  there are issues with a lot of jitter and associated latency increase, this does differ slightly depending on the destination traced, but once again i have slow throughput this time it's also affecting multi threaded downloads and the upstream isn't  good either, here's what i mean and this
How long do plusnet expect people to tolerate this almost nightly interruption in service with no proper answers and no solution in sight ?
And other observations that i have made  are  that the number of customers connected to  a gateway makes no  real difference, to throughput,ect, the only differences in gateways  i have found is that the newer BNG's do provide a small increase in headline throughput over the older gateways, the latency difference could be down to the gateway config, or the central ( 10 or 11) or the particular endpoint so can sometimes appear to change

[#3]

i did hopping around at 7pm ish after the MS windows 10 live stream which was slow as hell and buffered alot...but I think that was due to MS being bad at streaming, maybe running it off a windows server
Anyways
I hopped a lot and i didnt see much slowdown...the throughput on some of the gateways was up and down but i'm currently on ptw-bng02 seeing pretty stable speeds
I tend to avoid the pcl gateways because they're usually the ones I have trouble with 😐
Pic below of 10pm when i remembered to shut off the stream I was watching
Actually I figured I'd post all 3 today! Or more...
The 5pm and 6pm are on a pcl gateway, the rest are on ptw-bng02
The reason I switched was the pcl gateway was starting to look erratic...
In hindsight I should have stayed on it and looked at it throughout the night
But at the peak times at 7pm, 8pm, 9pm, 10pm...you can see it's very very stable throughput with pretty much no divergence between the TBB test and HTTP.

[#4]

Interesting why only the 2 speedtests carried out pre peak time @16:58 and again @18:00 have the ISP tag displayed on them all the others following those dont have the identity of the ISP on them they are marked unknown , every time that i have used the think broadband speedtester  it has always identified  my ISP /IPblock correctly , so those tests to me are not conclusive as without evidence of which isp's connection used  they could belong to any other isp couldn't they  the only possible cause i could see is that you have a dynamic ip that isn't registered properly with plusnet  
Anyway as predicted as we approach 11pm  throughput is starting to normalize the upstream is operating as intended again the downstream is getting somewhere near
last night ping grap shows the packet loss stop when i swapped gateways , and tonight's laggy peak time ping monitor And the usage graphs IMO tally with the slow downs ect what i and others are seeing every night most of which used to be limited to 1-2 days a week, now its present on a daily basis http://i.imgur.com/5kubpVD.jpg 

[#5]

Quote from: deathtrap

Interesting why only the 2 speedtests carried out pre peak time @16:58 and again @18:00 have the ISP tag displayed on them all the others following those dont have the identity of the ISP on them they are marked unknown , every time that i have used the think broadband speedtester  it has always identified  my ISP /IPblock correctly , so those tests to me are not conclusive as without evidence of which isp's connection used  they could belong to any other isp couldn't they

There's a good reason for that...
It's definitely Plusnet even if some applications think I'm american
Also if you don't trust mine then you can't trust anybodys TBB speed test since it's easy to change what ISP it displays via the drop down menu.

[#6]

I didn't say or imply  that i didn't trust you, it does look a bit suspect though you got to admit,So yet another thing that plusnet need to do  get their new IP's registered , i suppose there is one or 2 advantages  though if you do a lot of P2P it would probably fool the likes of ACS law

[#7]

I have just been booted,

Jan 22 01:14:05 TomatoAC66 daemon.info pppd[23111]: No response to 6 echo-requests
Jan 22 01:14:05 TomatoAC66 daemon.notice pppd[23111]: Serial link appears to be disconnected.
Jan 22 01:14:05 TomatoAC66 daemon.info pppd[23111]: Connect time 4234.8 minutes.
Jan 22 01:14:05 TomatoAC66 daemon.info pppd[23111]: Sent 1013133175 bytes, received 2428276140 bytes.
Jan 22 01:14:11 TomatoAC66 daemon.notice pppd[23111]: Connection terminated.
Jan 22 01:14:11 TomatoAC66 daemon.notice pppd[23111]: Modem hangup
Jan 22 01:14:15 TomatoAC66 user.info redial[23112]: WAN down. Reconnecting...

If I am on a slow endpoint now I will stay on it until friday evening and report the slow speeds.

[#8]

Ooook yeah so there really is gateway issues.
Proper gateway issues. The first pic is on ptn-bng01...good speeds, mostly stable, no divergence between HTTP and TBB tests.
The second pic is on pcl-bng02 at 12.46am. That one is all over the show

[#9]

Hi all,
I'm still passing this information on to the team. They're investigating what could be causing the slowdowns and are collating the data you're providing as part of this.

[#10]

Is there an issue with the bng 02 gateway then  ?
I posted yesterday about the increase of users but this was expected due to the number of endpoints or something.
http://www.plus.net/support/service/network_performance/broadband_bandwidth_usage.shtml

[#11]

@Justin7
Nothing we're aware of.
Update to the others in this thread:
We've continued looking at this from multiple angles and our testing is pointing towards SVLAN usage. We can't see anything wrong with our network, endpoints etc.
deathtrap - AMBER
goldenfibre - AMBER
nbk - AMBER
chrocoluk - AMBER
tonyAtplus - AMBER
Melancholie - Resolved(ish) - http://community.plus.net/forum/index.php/topic,135389.msg1191813.html#msg1191813
The SVLAN reports updated earlier this week and we've been through KBD tests with the above results showing.
And we've just checked a secondary data source and they *aren't* all amber. We're still investigating anyway!

[#12]

I think the fact that all those users can re-connect and get normal speeds again is showing it might be BTW backhaul after the SVLANs.
AAISP certainly have a few cases for this on-going at the moment in the South West and parts of London.

[#13]

We are seeing some levels of congestion in the evening on the following exchanges: BT COWBRIDGE, BT MORRISTON, BT WEST (Bristol area), BT CARDIFF EMPIRE, BT THORNBURY, BT EASTON, BT WINTERBOURNE, BT FISHPONDS, BT LLANTWIT MAJOR. These have been reported to BT and they are currently investigating.
I live now where near those locations, and those issue have been ongoing since 8/1/15 https://aastatus.net/2065

[#14]

I did say in the South West and parts of London...

[#15]

Looks like either Plusnet are to blame or the issues people are complaining about are such that they aren't registering on Andrews and Arnold's telemetry, which isn't beyond the realms of possibility.