High packet loss and other DNS issues
I remember before with Plusnet, we used to have those gateway graphs showing us congestion on each gateway.
Now that we've moved to a new network, I simply think that these packet loss issues is an problem with congestion. With the old network, I remember the links like PCL-BNG01 were the culprits around 6pm (evening) causing packet loss and bandwidth issues. The gateway's were less sticky and a router disconnect would have changed your gateway easily, but not any more 
Congestion somewhere along the path. Congestion occurs when the amount of traffic destined for a particular link exceeds the capacity of that link. The switch or router will buffer some packets, but eventually packets must be discarded. Congestion is most common at the entrance to a wide area network -- where the high capacity LAN meets a lower capacity access link -- and at the exit from the wide area network -- where the service provider's large network meets the lower capacity access link.
Essentially that link is overloaded and that everyone's packets are placed on a queue in a FIFO I/O. We no longer have any statistics about which gateway is at maximum capacity, but this is the only logical reason (which makes sense if Plusnet are silently acknowledging any issues in their official service status pages).
However, it is possible that it is this other reason
Loss due to bit errors corrupting the packet contents. With modern transmission technology this is less likely, but some links may have high bit error rates. Corrupted packets may be detected at Layer 2 and the packets will be discarded in the network.
Also this sounds very familiar to the issues I had previously when visiting a web server hosted on Virgin Media. There is no longer any issues with that server, but just a thought. If I were to ping the VM web server and see if packet loss is occurring, I believe one path on the gateway is causing this issue.
https://www.plus.net/features/network/
Unlike many ISPs, PlusNet own and manage their broadband platform. This means problems or configuration requirements can be quickly reacted to.
Not quite sure that this issue has been dealt with quickly or even reacted to.
After a IP address change at around 4:30 AM - I'm now in the 146.198.150.x range - I seem to be having the horrendous packet losses again, after spending the last few days with only 2-3%. Though, to be fair, I'm showing much lower packet loss (0-5%) when I ping pingbox1.thinkbroadband.com than it seems to be getting to me (40%+).
Another brilliant day.
Seems like bank holiday has started more longer periods of packet loss.
I can't see how this new network should be trialled out wide spread when we have serious issues with packet loss. Until, this is solved you guys should have just reverted mentioned users comments of packet loss onto the old network until a solution is found.
I think that there is enough data gathered here on this thread the past 4 weeks to diagnose the issue. Likely to be a routing issue or congestion ( seeing that this only happens after the early mornings).
@Riza wrote:
I remember before with Plusnet, we used to have those gateway graphs showing us congestion on each gateway.
Now that we've moved to a new network, I simply think that these packet loss issues is an problem with congestion. With the old network, I remember the links like PCL-BNG01 were the culprits around 6pm (evening) causing packet loss and bandwidth issues. The gateway's were less sticky and a router disconnect would have changed your gateway easily, but not any more
Now that I'm seeing this on my own circuit at home, I can see that the loss seems to occur at points past what you'd consider the 'gateways' to be on the new network (the 'IR' hops in an outbound traceroute). I agree, the symptoms do look like congestion, however there should be plenty of capacity across the parts of the network in question.
@scottrc wrote:
It's been 4 weeks now with me this problem, so I think it's fair to start asking for official answers. So could any plusnet staff please answer these. Are there currently any engineers out and about looking into this problem, or are you still only gathering data from customers?
Yes, there are engineers looking into the problem and a number of physical changes/maintenance tasks have been carried out over recent weeks. Unfortunately, and as you're only too aware, the problem remains unresolved 
@Riza wrote:
I can't see how this new network should be trialled out wide spread when we have serious issues with packet loss. Until, this is solved you guys should have just reverted mentioned users comments of packet loss onto the old network until a solution is found.
This isn't really an option. Moving a user back to the 'old' network takes weeks and eats into the capacity we've set aside to migrate customers the other way (from old to new). Assigning a static IP does push the connection back through the 'old' L2TP network, however because of the way IP resource has been reassigned, it's not feasible to do this for all those on the new network either.
We are working hard to get to the bottom of this and we appreciate both the patience, and continued observations from you guys.
Hopefully have something more substantial to report this week.
