Plusnet outage

[#1]

Ok, I seem to be back up. 
Not sure in the end how much downtime was Plusnet and what was down to me. I'd restored my config from a backup after my pfSense upgrade went badly wrong and had forgotten that I changed the Plusnet password in the intervening period. 
No modem reset necessary.
Steve

[#2]

mine is showing DNS as 212.159.6.9 and 212.159.6.10 which are issue problem. Changing to this: Primary DNS: 217.169.20.20 & Secondary DNS: 217.169.20.21 solved the issues problem.

[#3]

Quote from: Oldjim

For information Andrews and Arnold are not showing any problems so I suspect that it is a Plusnet problem rather that BTw

That doesn't always follow as AAISP might have their LNS's in a different connection connected to a different POP to Plusnet.

[#4]

https://www.whatsmydns.net/dns/uk I changed mine to AAISP DNS and no issues problem!

[#5]

Thank gawd for that  stephenw10
@Jim http://community.plus.net/forum/index.php/topic,131222.msg1144348.html#msg1144348 suggest don't merge either.

[#6]

@goldenfibre
212.159.6.9,  212.159.6.10,  212.159.13.49 & 212.159.13.50 are all behind a load balancer so it shouldn't really make too much difference. Ifyou are having trouble with one, then you may with the others, but then it could be your current routing causing the problem or anything else.

[#7]

AAISP also use the TalkTalk backhaul - I have no idea on the customer split between 21CN and TT. I assumed TT was cheaper than using the BT Wholesale 21CN network, but apparently it's similar http://aa.net.uk/kb-broadband-ttbt.html
I'm not sure if PN will explain exactly what the problem is/was - but if you look at the POSI (Point of Service Interconnect - the link between the 21CN network and PN's datacentres) graphs here http://www.plus.net/support/service/network_performance/broadband_bandwidth_usage.shtml, you will see the outage only affected 20/21CN users connected via Colindale. This to me would indicate an issue with the interconnect, which supplied by BT Wholesale. But it's all a bit of a guess by me.

[#8]

A bit uncertain there, they could be graphing bugs, and bearing in mind some people couldn't get back on for hours, that's not reflected in those graphs. And the failure occurred long before that time.

[#9]

This is true - but I think those bandwidth graphs are fairly accurate. I would hazard a guess that a network device was failing and then completely failed just after 15:00.

[#10]

Just for giggles I was disconnected at 13.39 and reconnected at:

Aug 29 16:22:51  rc_service: ntp 715:notify_rc restart_upnp  WAN Connection: WAN was restored.

[#11]

Not very giggly really    Interesting the time you went down launton.
Andy, If you look at the collated users online, something died not long after midday maybe nearer 1300. It looks as though things were going down fromthen on. Service Status was posted at 1338 and then a lot more went just before 1430 or thereabouts.

[#12]

You can see in my signature a rough estimate of when things went down for me. Can't give you exacts because the router locked up when it went down for some reason and had to be rebooted.

[#13]

The users online graph looks like lots of users dropped off ptn-bng01, ptw-bng01 and pcl-bng02, but after the first drop around 13:00, lots of users quickly got reconnected to the ag gateways instead of the newer bng ones.

[#14]

I was connected to ptw-bng01 when it died at 15.00 & then reconnected at 16.15 to ptw-ag03. At 18.00 I did a soft reset & connected to pcl-bng02 which is one of my preferred gateways.

[#15]

Quote from: Xreyuk

...........Can't give you exacts because the router locked up when it went down for some reason and had to be rebooted.

This can happen, others and myself have experienced this with a variety of modem/routers, it seems it just depends on the way things fail
See my suggested advice in replies #71 & #72 especially if on FTTC but also this if on ADSL more so if on 20CN, and on a longer line. On a short line you can get away with more without necessarily affecting the profile.