Intermittent And Substantial Latency & Packet Loss

[#1]

Reported a fault earlier today via chat about massive latency and packet loss which rather unhelpfully resolved itself briefly at the time. Had a few hours good use out of the service and then the problem returned just after 10pm when customer services had already gone home unfortunately. Will report again first thing in the morning of course. 
Not an understatement to say that my connection is unusable in this state. Latency is averaging >1000ms with roughly 20-30% packet loss. Attached a couple of comparative ping traces showing the issue which shows the 1st hop after my local router (entry point to plusnets network) as being very unhappy indeed. I can think of a few explanations none of which involve my local infrastructure (which passed all the diagnostics).
Anyone else having similar issues?

[#2]

Still continues this morning. Currently waiting on the phone to talk to customer services.
Took the opportunity to run the BT Wholesale speedtests last night. Failed to complete on 3 occasions recording zero values in upstream speed and ping latency.
Any comments would be handy. A few days holiday from work and my internet connection is completely buggered! Wonderful!

[#3]

I've tested your line and there's no obvious issues being reported. The line is short and is connecting at a good speed (although lower than I'd expect for the length of line). The phone line is testing fine and there are no errors showing.
Are you testing wired or wireless? Is there any noise on the phone line?

[#4]

Ping trace is done from my desktop which is wired but the same issue replicates with devices using wifi such as my PS4 downstairs. Though I can't ping trace from there I can't get a stable connection to a server when gaming.
No notable noise on the phone line either.
I'm certain it isn't the router as the diagnostics run fine and show a stable connection with no  errors but I will try and borrow one this afternoon just to rule it out completely. I'll update when done.

[#5]

I have subsequently checked the following:
1. Replaced router - no change.
2. Replaced ADSL cable from router to primary phone point - no change
3. Replaced CAT5 cable from desktop to router - no change
4. Tried secondary phone point with a different microfilter - no change
.... and I've practically had to dismantle my spare room to do it!
Attached a further ping trace to show the continuing issue below.
Almost 100% certain at this point that it isn't an issue with my internal infrastructure. Worth also noting that I had complete loss of internet somewhere along the way. I've only just managed to reconnect!
Please investigate further.

[#6]

Finally got the BT Speedtest to report something meaningful. Attached shows the issue - connection speed is fine but latency is huge! In 2 out of 3 cases a low or zero value was reported for upstream speed whilst downstream was relatively consistent suggesting that the issue is on the upstream side probably.
I've had a couple of issues like this before caused essentially by PSTN faults but in both cases the connection was very unstable and there was a huge amount of noise on the phone line. I've had one issue similar to this before with another provider where a router beyond their network was de-prioritising certain types of traffic leading to a very similar issue overall. The main difference there was that it only applied to a specific game (WoW) so as long as I wasn't playing that I didn't have any issues. If the issue had been with the ISP or somewhere between me and the ISP then this problem would look like what I'm now reporting. I'm guessing that Plusnet's infrastructure is behaving as it should and there are no traffic management policies in force either by policy or as a result of compensating for a failing router within the network?
Not sure I can be much more help on this one. Fair to say this is doubly frustrating as I have a few rare days to waste gaming and now can't because my connection has fallen over! Typical 

[#7]

Issue continues this morning. It appears to be cyclical - good for a while then bad for a while. Seems like a fairly regular pattern throughout the day too so far.
I'll keep checking it periodically and update.

[#8]

If your router supports WAN side ping response, establishing a TBB BQM would be useful.

[#9]

Thanks I'll check this out but to be honest I'm not really wanting to spend any more time on this. The problem continues and its relatively regular in its severity (it gets better and worse at similar times of day every day) which suggests that its 'manmade' rather than circumstantial or transient. This doesn't bode well for a quick fix. Despite speaking to customer support twice in the last week and logging a fault at their request I've heard nothing more on this subject from plus.net other than what has been posted here by Chris and we are now well past the prescribed 48-72 hour contact window. I'll leave it until tomorrow (being that its the first proper working day after the Xmas break) but if I hear nothing and the problem persists I will have to consider going elsewhere for my phone and broadband. This is a shame because up until now both my connection and plus.net customer service have been very good.
mav:quote

[#10]

If - as is most likely - the problem is related to the line between your home and the exchange, changing ISPs is NOT going to fix the problem.  Self help / education is your best bet on understanding the problem.  Intermittent issues are the hardest to locate and resolve, as it is enevitable that the problem will not be present when inspected.  Self sought evidence is essential to progress such matters, if for no reason other than to identify 'failure' time slots.
If this is REIN caused, then it is an environmental problem, potentially nothing to do with BTOR, let alone PlusNet.
This is likely not what you want to hear, but it is honest advice!

[#11]

Don't mind an honest response. I appreciate the insight to be honest - thanks!
I do have to disagree with you on REIN though. I live alone which means I would be in complete control of any electrical appliance that is turned on or off in my flat. I have added no new equipment to my setup since before the problem started and have taken steps to minimise the potential by running with almost nothing on at all except that which is necessary to run my PC. Moreover, the connection is completely stable with no drop outs. Using the BT speedtest the downstream line speed is also stable but the upstream is all over the place. REIN would more than likely affect both equally. SNR margin is also completely stable at 6 and has been since the connection established earlier today (following a reboot of the router at my end). Prior to the recent investigations I've been doing in conjunction with Plusnet the connection had been up for months without interuption. I will however revisit this one as you've mentioned it. Doesn't hurt to check everything twice.
Self help and education is something I am used to having had a couple of intermittent faults like this one - 2 x PSTN faults and one overloaded router on the way to the servers of a well known creator of MMOs which was dumping traffic at peak times because of excessive load to be exact. As you say intermittent faults are not easy to diagnose and generally you have to rule out the things that can be ruled out and see what you are left with. Changing ISP may not fix the problem but it rules out bottlenecking at the gateways or intermediate infrastructure - this is something I have also encountered in the past and was forced to switch as a result. To be honest my reason for moving at this point would be as much down to customer service as anything else. I found out today that the fault I reported on the 30th via chat in fact wasn't captured (despite having the confirmation from the customer service rep that it was) meaning I had to resubmit today and wait another 48-72hrs. Since there are things that I can't check without the ISPs cooperation it would be kind of nice if they wanted to engage a little bit more in solving the issue. Especially as I've spent a huge amount of time already working out what it isn't (or most likely isn't).
Still unclear about the capabilities of my router with regard to WAN side ping response. Its a Draytek Vigor 2710 and has proved to be pretty capable in most respects. I'll keep looking into it.

[#12]

Hi,
REIN can come from anywhere.  I've seem passing electric trains cause massive SNRM spikes, sometimes enough to drop the line.  Not all trains exhibited the problem and not all that did broke the service.  From the detailed monitoring I could discern the direction the train was travelling in.
Note that I said resolution of REIN is likely to be outside of BTOR's control; as you indicate, if it is from within your premises, you can have control over it.  Often is not a matter of something new, but rather something old which has degraded, for example switched mode (current) power supplies.
First though we need to profile the problem. Getting the TBB BQM established and running router stats would be very helpful.  There are other tools about there as well, for example xDSL stats.  The thread below which offers advice on configuring router stats is a good illustration of the multiplicity of factors which can impact a service.  That user had nearly every issue going...

@CRT,
Can you please post a VR plot?  Also is there any insight as to why the why the line is running slower than expected?

[#13]

Struggling to get my Vigor 2710 to connect to RouterStats but reading through the help and forums this might be because of the way you login - embedded in a webpage rather than as a popup. I'll have another play with this this morning. Failing that I've got an older Vigor 2800 that I used to test the connection which does look to be supported at least by RouterStats Lite. If I recall this does generate a pop up for login. Either way I'll get something sorted today so we can start monitoring line stats.
I pretty sure that both the Vigor 2710 and the 2800 will support WAN side ping response. Reading through the manual there is just a box that needs to be checked that allows the router to respond to to ping requests. Once I've sorted out RouterStats I'll enable the ping response from the router and set up a TBB BQM with a provider.
Lets hope it helps.

[#14]

You might find that you can bend the telnet options of router stats to access the router.
One of the other tools might be worth looking at as well.

[#15]

I found a way to monitor SNR Margin and Downstream Speed using SNMP. Bit of a faff but I couldn't get any of the tools you suggested to work with my router (probably me).
How long should I run it for?