Peak time latency spikes

[#1]

@ Jaggies
Thanks - other users on my exchange got disconnected too, so is part of their upgrade.

[#2]

Jaggies, AndyH, that Maintenance announcement is general work on BT's network not local exchange specific stuff. You might find this link more useful for local stuff. The problem is when you have linked numbering, the info isn't always specific enough the identify which exchange(s) are affected. You can sometimes get (after the event) data here.

[#3]

Although the PEW notification mentions specific locations, it also points out that

Quote

due to the nature of the broadband network these outages may affect customers nationwide. There are also a number of smaller exchange works most nights of the week that may cause smaller outages at individual exchanges.

The timing (between midnight and 6:00am) and length of the disconnection (less than 30 minutes) would tend to make it much more likely to be the PEW that BTW is and has been running every week-day night for several months now.

[#4]

But that is in effect what I said. In other words it's not specifically the local exchange upgrade work that AndyH was interested in.
Anyway this is now off-topic.

[#5]

Last night on pcl-ag04 was a big spike from pretty much dead on 9pm -

With respect to BT engineering work, my exchange is also a core node in their 21CN network (all other areas in that list are other core nodes too). We believe they were upgrading capacity from these nodes due to peak time congestion.

[#6]

AndyH:  That was this spike problem again.  I was online with the engineer working to sort it.  How did your broadband experience suffer? 

[#7]

To be honest...if I had not checked the graph, I would have never known there was a minor issue.
We didn't notice any web page slow downs, or any problems streaming from Sky Go/iPlayer.
Edit: If you compare this to the recent issues with packet loss (which was a problem with BT's network and nothing to do with PN), then you could tell something was not right - pages not loading/having to continually press refresh, buffering issues, downloads halting etc. But with the spike problem, I can't tell something is wrong.

[#8]

FYI, we're doing some more work tomorrow morning to try and help alleviate the problem:
http://usertools.plus.net/status/archive/1360258611.htm

[#9]

no sign of it yet on my graph or others but speed has halved and watching any streams is painful
ptw-ag02

[#10]

Can you get some speed tests, and perhaps the iplayer speed test?  then reconnect to a different gateway or 2 and repeat for us?

[#11]

speed test history for today etc, as for the BBC speed test never found it accurate, always gives my sync speed

[#12]

Saturday was really bad for me on several gateways - see http://community.plus.net/forum/index.php/topic,111842.0.html although I guess I may have had a different problem.
It's really becoming quite a pain now, quite a contrast to my normal very reliable connection. To be honest I'm surprised that this doesn't have greater visibility - shouldn't the broadband service status reflect this until it's fixed.
Bob

[#13]

Bad speeds and latency again this morning    I hopped gateways multiple times until I eventually got one that is working!  Here's all the gateways I tried and latency to them

817 ms  lo0-central10.pcl-ag01.plus.net [195.166.128.182]
750 ms  lo0-central10.pcl-ag04.plus.net [195.166.128.185]
694 ms  lo0-central10.pcl-ag07.plus.net [195.166.128.188]
754 ms  lo0-central10.pcl-ag08.plus.net [195.166.128.189]
841 ms  lo0-central10.ptn-ag01.plus.net [195.166.128.190]
954 ms  lo0-central10.ptn-ag02.plus.net [195.166.128.191]
35 ms   lo0-central10.ptn-ag04.plus.net [195.166.128.193]

Strange thing on all the gateways that I had high latency it wasn't immediately after connection.  A few seconds after connecting the latency started to climb before leveling off

Request timed out.
Request timed out.
Request timed out.
Request timed out.
Reply from 8.8.8.8: bytes=32 time=44ms TTL=47
Reply from 8.8.8.8: bytes=32 time=41ms TTL=47
Reply from 8.8.8.8: bytes=32 time=42ms TTL=47
Request timed out.
Reply from 8.8.8.8: bytes=32 time=44ms TTL=47
Reply from 8.8.8.8: bytes=32 time=43ms TTL=47
Reply from 8.8.8.8: bytes=32 time=42ms TTL=47
Reply from 8.8.8.8: bytes=32 time=42ms TTL=47
Reply from 8.8.8.8: bytes=32 time=42ms TTL=47
Reply from 8.8.8.8: bytes=32 time=43ms TTL=47
Reply from 8.8.8.8: bytes=32 time=42ms TTL=47
Reply from 8.8.8.8: bytes=32 time=171ms TTL=47
Reply from 8.8.8.8: bytes=32 time=235ms TTL=47
Reply from 8.8.8.8: bytes=32 time=278ms TTL=47
Reply from 8.8.8.8: bytes=32 time=342ms TTL=47
Reply from 8.8.8.8: bytes=32 time=383ms TTL=47
Reply from 8.8.8.8: bytes=32 time=432ms TTL=47
Reply from 8.8.8.8: bytes=32 time=466ms TTL=47
Reply from 8.8.8.8: bytes=32 time=496ms TTL=47
Reply from 8.8.8.8: bytes=32 time=499ms TTL=47
Reply from 8.8.8.8: bytes=32 time=548ms TTL=47
Reply from 8.8.8.8: bytes=32 time=576ms TTL=47
Reply from 8.8.8.8: bytes=32 time=594ms TTL=47
Reply from 8.8.8.8: bytes=32 time=636ms TTL=47
Reply from 8.8.8.8: bytes=32 time=654ms TTL=47
Reply from 8.8.8.8: bytes=32 time=655ms TTL=47
Reply from 8.8.8.8: bytes=32 time=704ms TTL=47
Reply from 8.8.8.8: bytes=32 time=714ms TTL=47
Reply from 8.8.8.8: bytes=32 time=754ms TTL=47
Reply from 8.8.8.8: bytes=32 time=775ms TTL=47
Reply from 8.8.8.8: bytes=32 time=782ms TTL=47
Reply from 8.8.8.8: bytes=32 time=784ms TTL=47
Reply from 8.8.8.8: bytes=32 time=824ms TTL=47
Reply from 8.8.8.8: bytes=32 time=839ms TTL=47
Reply from 8.8.8.8: bytes=32 time=881ms TTL=47
Reply from 8.8.8.8: bytes=32 time=890ms TTL=47
Reply from 8.8.8.8: bytes=32 time=892ms TTL=47
Reply from 8.8.8.8: bytes=32 time=894ms TTL=47
Reply from 8.8.8.8: bytes=32 time=933ms TTL=47
Reply from 8.8.8.8: bytes=32 time=921ms TTL=47
Reply from 8.8.8.8: bytes=32 time=467ms TTL=47
Reply from 8.8.8.8: bytes=32 time=514ms TTL=47
Reply from 8.8.8.8: bytes=32 time=546ms TTL=47
Reply from 8.8.8.8: bytes=32 time=583ms TTL=47
Reply from 8.8.8.8: bytes=32 time=610ms TTL=47
Reply from 8.8.8.8: bytes=32 time=648ms TTL=47
Reply from 8.8.8.8: bytes=32 time=650ms TTL=47

Sticking with ptn-ag04 for now - I hope I don't have to reboot/reconnect too soon!

[#14]

I hope you aren't Rebooting your modem/router every time just to Gateway hop. If so, you should login to your modem/router interface and use the Disconnect/Connect buttons to drop and reconnect the PPP session.This will avoid loss of sync each time which will be seen by the Exchange DLM as a dropping connection.

[#15]

Only disconnecting the PPP session, not resyncing with the exchange 
I should also still be opted out of the DLM as per a trial some time ago, but  no point in risking it with too many reboots.