Packet Loss on Plusnet Servers
@DS wrote:
137 and 141.hiper04 receiving packet loss of like 70%
@MasterOfReality Can you please speak to @Kelly about this.
I've lost count how many times I've raised 137 hiper and 141 hiper
Kelly did post that he'd been heard chatter on 141 hiper around 3 weeks ago now, he was going to check, but never came back.
What was said about 141 hiper - it'll solve my issues too
(he posted on an xbox topic)
@Gandalf So why oh why do I sometimes see this?
3 * * 54 ms 141.hiper04.sheff.dial.plus.net.uk [195.166.143.141]
4 12 ms 10 ms 10 ms 140.hiper04.sheff.dial.plus.net.uk [195.166.143.140](hop 3 either times out completely, shows like this, or is there - Kelly was going to check as he wasn't sure why it's either there in full, part there, or missing completely)
The x.hiper04.sheff.dial.plus.net.uk hops are routers. The primary function of these routers is to route traffic, not respond to ICMP requests, especially when they are busy. The fact that the ping plotter results show no packet loss at the final hop and low latency, means that we can ignore the packet loss observed on the x.hiper04.sheff.dial.plus.net.uk hops, we've actually been to check each of our interfaces and they're each fine, graphing too looks normal.
There's no problem 
@JonoH Wrote:
The x.hiper04.sheff.dial.plus.net.uk hops are routers. The primary function of these routers is to route traffic, not respond to ICMP requests, especially when they are busy. The fact that the ping plotter results show no packet loss at the final hop and low latency, means that we can ignore the packet loss observed on the x.hiper04.sheff.dial.plus.net.uk hops, we've actually been to check each of our interfaces and they're each fine, graphing too looks normal.
There's no problem
Thanks for that detailed reply Jono :). I'm still confused, but that's probably just me...;)
So, being on a dynamic IP, this clearly routes through your 'busy' routers/interfaces then?
As your 'busy' routers/interfaces are not used when on a static IP?
So is going static the only way to prevent packet loss within your network?
But why only hop 3? As hop 4 is always fine. They are both hiper.04 routers are they not? Is there a congestion issue at all routers/interfaces sitting at hop 3, as repeat tests, to different 04.hiper routers, differ so much on my connection? (day or night, others on my connection, or just me)
Could the odd, ignored by you packet loss, be a cause of my connection issues, dropped from gaming sessions, forum pages not loading quickly, failure to post on these forums (rare but happens), forum glitches, lagging games, issues with Discord etc. I know you said there is no problem, I just wished I know where my problem was and how I can fix it!
I don't see any missing results on hop3 on other connections (either other posters on a static IP) or testing my laptop on a non Plusnet connection. These are a combination of external static IPs and dynamic IP's. I only see it on Plusnet:(
I just kinda thought it would be either there all the time, or not there, not how I see it, with times missing.
Anyhoo, the reason for the delay is actually getting to the forum servers, it's been, well:
C:\WINDOWS\system32>tracert community.plus.net
Tracing route to d14xs8zr41zt1m.cloudfront.net [54.192.33.23]
over a maximum of 30 hops:
1 1 ms 1 ms 1 ms PNHUB1 [172.16.0.1]
2 * * * Request timed out.
3 16 ms 17 ms 17 ms 141.hiper04.sheff.dial.plus.net.uk [195.166.143.141]
4 17 ms 24 ms 17 ms 140.hiper04.sheff.dial.plus.net.uk [195.166.143.140]
5 18 ms 18 ms 18 ms 195.99.125.138
6 18 ms 17 ms 17 ms 109.159.252.140
7 17 ms 17 ms 17 ms 213.137.183.34
8 18 ms 18 ms 18 ms ldn-b9-link.telia.net [213.248.97.50]
9 19 ms 18 ms 19 ms ldn-bb3-link.telia.net [62.115.117.6]
10 18 ms 18 ms 17 ms ldn-b7-link.telia.net [62.115.138.151]
11 20 ms 17 ms 18 ms a100row-ic-304715-ldn-b3.c.telia.net [213.248.101.150]
12 * * * Request timed out.
13 * * * Request timed out.
14 * * * Request timed out.
15 * * * Request timed out.
16 * * * Request timed out.
17 * * * Request timed out.
18 * * * Request timed out.
19 * * * Request timed out.
20 * * * Request timed out.
21 * * * Request timed out.
22 * * * Request timed out.
23 23 ms 23 ms 22 ms server-54-192-33-23.man50.r.cloudfront.net [54.192.33.23]
And
C:\WINDOWS\system32>tracert community.plus.net
Tracing route to d14xs8zr41zt1m.cloudfront.net [54.192.33.23]
over a maximum of 30 hops:
1 1 ms 1 ms 1 ms PNHUB1 [172.16.0.1]
2 * * * Request timed out.
3 * * 17 ms 141.hiper04.sheff.dial.plus.net.uk [195.166.143.141]
4 17 ms 17 ms 17 ms 140.hiper04.sheff.dial.plus.net.uk [195.166.143.140]
5 18 ms 17 ms 17 ms 195.99.125.138
6 18 ms 18 ms 17 ms 109.159.252.140
7 17 ms 17 ms 16 ms 213.137.183.34
8 18 ms 18 ms 20 ms ldn-b9-link.telia.net [213.248.97.50]
9 19 ms 18 ms 19 ms ldn-bb3-link.telia.net [62.115.117.6]
10 18 ms 18 ms 18 ms ldn-b7-link.telia.net [62.115.138.151]
11 17 ms 17 ms 17 ms a100row-ic-304715-ldn-b3.c.telia.net [213.248.101.150]
12 * * * Request timed out.
13 * * * Request timed out.
14 * * * Request timed out.
15 * * * Request timed out.
16 * * * Request timed out.
17 * * * Request timed out.
18 * * * Request timed out.
19 * * * Request timed out.
20 * * * Request timed out.
21 * * * Request timed out.
22 * * * Request timed out.
23 23 ms 23 ms 24 ms server-54-192-33-23.man50.r.cloudfront.net [54.192.33.23]
Guess it's busy after the first tracert on hop 3 then
(but the remainder is outside of your control, just really at the point of why bother posting, it's not much fun atm)
So, being on a dynamic IP, this clearly routes through your 'busy' routers/interfaces then?
They're not always busy
Being on a dynamic IP runs through our network, previously people have referred to it as our "new" network but in fairness, almost all of our customers are routed this way now. A static IP address routes you through our network but in a different way, it more like how it used to be before the introduction of the "new" network and sometimes moving you onto a static IP can help us troubleshoot.
So is going static the only way to prevent packet loss within your network?
You're not getting packet loss, as we said earlier the fact that the final hop shows no packet loss and a low latency we know that there can't have been any loss in the previous hops.
(Thinking out loud here)
So, just to get my head around my clearly incorrect thought process.
Being on a dynamic IP routes it through some PN network routers that don't always respond, or are at times busy, and gaming is sometimes an issue?
And being on a static IP, routes it a different way, thus 'skipping' these PN network routers that don't always respond, or are at times busy, and gaming is sometimes not an issue?
But there's no issue with dynamic IP's or how they're routed?
I still don't get the time I put 2x XB's next to each other, same game, same session, 2x TV's, 1x PN connection yet both characters were not in the same place on each of the screens. Kinda glad I stopped gaming now
<goes to look at what Sky has to offer;), no guarantees I'm jumping just yet>
@DS wrote:
(Thinking out loud here)
So, just to get my head around my clearly incorrect thought process.
Being on a dynamic IP routes it through some PN network routers that don't always respond, or are at times busy, and gaming is sometimes an issue?
> PN network routers that don't always respond, or are at times busy,
Don't always respond to ICMP requests and that is normal. We don't have any evidence that suggests they are over busy. That doesn't mean there isn't an issue we haven't spotted yet though, obviously putting the static on fixes something.
And being on a static IP, routes it a different way, thus 'skipping' these PN network routers that don't always respond, or are at times busy, and gaming is sometimes not an issue?
But there's no issue with dynamic IP's or how they're routed?
The dynamic changes a couple of things, Firstly the traffic is in a tunnel which means you don't see all the devices it goes through which is why the traceroutes change significantly. Also, there are different cards in our equipment which handle it. Potentially there is something wrong here.
I still don't get the time I put 2x XB's next to each other, same game, same session, 2x TV's, 1x PN connection yet both characters were not in the same place on each of the screens. Kinda glad I stopped gaming now
Network games don't all work like that. There is loads and loads of netcode in them to make things seem seamless, but handle players with lots of different line characteristics/speeds/packetloss/latency. Your position on the screen relative to other players might not be an important thing for the game.
@FortKnox101 @JonoH will be in touch to see if we can arrange some testing with a Network engineer in ourside to see if we can pin this one down.
@Kelly So I'll ignore "ICMP requests" as they are normal:), it's the "busy" part that may or may not be an issue. And going static accepts the (normal) requests and diverts from the possible busy routers. So what can you suggest, not wanting to use TBB BQM monitor myself, that I can use to monitor this? I see PN staffers are suggesting to others to post up traceroute results, though now it does kinda seem to be normal ICMP request rejections. Is this advice a waste of every bodies time?
I'll admit it, I initially thought a traceroute showing timed out indicated a problem. I knew pinging our own PN router wouldn't work as our routers do not accept ICMP requests and didn't know your network routers did the same (even though some results show it responded fully).
I know 3rd party DDOS providers are or were being shut down, and not wanting a static IP (just asked my lad if he wants me to go static and his first words were about DDOSing too), one assumes my only recourse is a 3rd party VPN, if I wish to continue with my contract with the net of plus.
We don't have any evidence that suggests they are over busy. That doesn't mean there isn't an issue we haven't spotted yet though, obviously putting the static on fixes something.
Soooo, looking around at several new (and old) moans posts doesn't indicate a (packet loss) issue
? Some forum users didn't have issues and (using recent examples) they all of a sudden have issues? Yes understand this could be something completely outside of PN's responsibility.... but ya know;). Hope you find "something" before someone finds something else to use as their ISP
(not harping on, but my gaming issues only popped up when joining PN. We're putting some serious cash into fixing any issues that could be causing this at my end, hoping the investment is going to pay off)
Firstly the traffic is in a tunnel which means you don't see all the devices it goes through which is why the traceroutes change significantly. Also, there are different cards in our equipment which handle it. Potentially there is something wrong here.
I guess the tunnel is hop2 to hop4?
Openreach didn't want to tell me, although I really pushed (I contacted them and rose an official complaint direct to them). Is this related to the known OR issue <they want it to remain internal only, but insert which one here> of <a particular brand> of equipment in the cabinet (I'm sure they referenced it as a card) giving them a headache too? They did let me know they are actively swapping it out, but all they could say is it is known to be giving issues.
Network games don't all work like that. There is loads and loads of netcode in them to make things seem seamless, but handle players with lots of different line characteristics/speeds/packetloss/latency. Your position on the screen relative to other players might not be an important thing for the game.
Ok. But we only tried that when we were both being told from many other gamers that we were not next to them either. They were all fine, just us. They could interact with one another. It was like we were in a modded session, though we weren't. We soon sussed out why we has no chance of winning. But both getting kicked at the same time too? Router didn't drop and on separate devices...? Then our 'dead' characters performed normally without the others seeing any oddness.
But, it doesn't affect me directly now:). It will however once the final parts of the new gaming PC arrive and if the issues continue. I've even got a laptop on it's way mainly for Discord - Gaming PC doing the gaming (via Cat7) and Laptop doing Discord. If it works well, then we may take the laptop off and see how it goes too. If all's well then it'll be money well spent, if it carries on then clearly I won't be a happy bunny.
(nowt personal;))
@DS wrote:
@Kelly So I'll ignore "ICMP requests" as they are normal:), it's the "busy" part that may or may not be an issue. And going static accepts the (normal) requests and diverts from the possible busy routers. So what can you suggest, not wanting to use TBB BQM monitor myself, that I can use to monitor this? I see PN staffers are suggesting to others to post up traceroute results, though now it does kinda seem to be normal ICMP request rejections. Is this advice a waste of every bodies time?
If you see a jump in latency or packet loss at one hop, that continues through to the end of the trace, that shows a packet loss or latency issue. That would suggest the router is started at (or something hidden before it) is introducing the issue. With the above scenarios, you can see that after the hiper hops, you've got 100% of packets being responded to, and good latencies. That shows is that packets are passing through the hiper hops fine. Does that make sense?
I'll admit it, I initially thought a traceroute showing timed out indicated a problem. I knew pinging our own PN router wouldn't work as our routers do not accept ICMP requests and didn't know your network routers did the same (even though some results show it responded fully).
I know 3rd party DDOS providers are or were being shut down, and not wanting a static IP (just asked my lad if he wants me to go static and his first words were about DDOSing too), one assumes my only recourse is a 3rd party VPN, if I wish to continue with my contract with the net of plus.
You shouldn't need a static IP for things to work on the network. That suggests something, somewhere is amiss. It could potentially be out of our control though. I.e. a 3rd party on the internet treating the traffic differently because of the different IP range.
We don't have any evidence that suggests they are over busy. That doesn't mean there isn't an issue we haven't spotted yet though, obviously putting the static on fixes something.
Soooo, looking around at several new (and old)
moansposts doesn't indicate a (packet loss) issue(not harping on, but my gaming issues only popped up when joining PN. We're putting some serious cash into fixing any issues that could be causing this at my end, hoping the investment is going to pay off)
We've had issues in the past which look like this and were something broken on our end. Which posts are you looking at? I'm not combing the forums at the minute and rely on the Forum people giving me a nudge
Firstly the traffic is in a tunnel which means you don't see all the devices it goes through which is why the traceroutes change significantly. Also, there are different cards in our equipment which handle it. Potentially there is something wrong here.
I guess the tunnel is hop2 to hop4?
It's from your router to our equipment, with all the openreach equipment carrying the tunnel.
Openreach didn't want to tell me, although I really pushed (I contacted them and rose an official complaint direct to them). Is this related to the known OR issue <they want it to remain internal only, but insert which one here> of <a particular brand> of equipment in the cabinet (I'm sure they referenced it as a card) giving them a headache too? They did let me know they are actively swapping it out, but all they could say is it is known to be giving issues.
Network games don't all work like that. There is loads and loads of netcode in them to make things seem seamless, but handle players with lots of different line characteristics/speeds/packetloss/latency. Your position on the screen relative to other players might not be an important thing for the game.
Ok. But we only tried that when we were both being told from many other gamers that we were not next to them either. They were all fine, just us. They could interact with one another. It was like we were in a modded session, though we weren't. We soon sussed out why we has no chance of winning. But both getting kicked at the same time too? Router didn't drop and on separate devices...? Then our 'dead' characters performed normally without the others seeing any oddness.
But, it doesn't affect me directly now:). It will however once the final parts of the new gaming PC arrive and if the issues continue. I've even got a laptop on it's way mainly for Discord - Gaming PC doing the gaming (via Cat7) and Laptop doing Discord. If it works well, then we may take the laptop off and see how it goes too. If all's well then it'll be money well spent, if it carries on then clearly I won't be a happy bunny.
(nowt personal;))
Do you have an open fault for your issue then? I've not see what's affecting your line. Does it act like the issues in this thread?
Yes, i'm having much better performance now!
