Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ronski]

HI Ixel, thanks for the suggestion, but I have changed the face plate, only thing left that I can change now is the modem cable it's self.

I wonder if the problem I experienced last night (which is the second time it's happened) is hardware related as rebooting the modem cures it, and as I have changed the modem I think it must me something in the cabinet. 

[SecTSys]

that graph looks similar to a test that i performed recently with a BT Engineer over at mine - i was getting that jagged strut on my download speeds. - in the end it was discovered that the fault was with the dslam itself requiring me to have a "lift and shift" - I can't explain what that is properly but something to do with the DSLAM i believe. - even if there are "acceptable speeds" with your service - have them run a line check anyway. i am sure you will find the fault to be external rather than internal.

Personally the speed tester and network diagnostics tools i use are here http://www.measurementlab.net/run-ndt and once run click the advanced tab option for all the detailed network diagnostics info

[Ronski]

I am starting to think it's a fault in the DSLAM, tried your test and got the following

Code: [Select]

Your test results
Summary Details Advanced
Your system: Windows 7 version 6.1
Java version: 1.7.0_07 (x86)

TCP receive window: 261360 current, 261360 maximum
3.242E-5 packets lost during test
Round trip time: 29 msec (minimum), 88 msec (maximum), 44.91 msec (average)
Jitter: 59 msec
0 seconds spend waiting following a timeout
TCP time-out counter: 251
185 selective acknowledgement packets received

No duplex mismatch condition was detected.
The test did not detect a cable fault.
No network congestion was detected.
No network address translation appliance was detected.

0.7908% of the time was not spent in a receiver limited or sender limited state.
20.4% of the time the connection is limited by the client machine's receive buffer.
Optimal receive buffer: 267632640 bytes
0 duplicate ACKs set

Code: [Select]

WEB100 Kernel Variables: Client: localhost/127.0.0.1 CurMSS: 1452 X_Rcvbuf: 87380 X_Sndbuf: 676216 AckPktsIn: 15426 AckPktsOut: 0 BytesRetrans: 2904 CongAvoid: 11750 CongestionOverCount: 0 CongestionSignals: 1 CountRTT: 15241 CurCwnd: 258456 CurRTO: 251 CurRwinRcvd: 261360 CurRwinSent: 5888 CurSsthresh: 130680 DSACKDups: 0 DataBytesIn: 0 DataBytesOut: 45401600 DataPktsIn: 0 DataPktsOut: 30845 DupAcksIn: 183 ECNEnabled: 0 FastRetran: 1 MaxCwnd: 262812 MaxMSS: 1452 MaxRTO: 288 MaxRTT: 88 MaxRwinRcvd: 261360 MaxRwinSent: 5888 MaxSsthresh: 130680 MinMSS: 1452 MinRTO: 229 MinRTT: 29 MinRwinRcvd: 45264 MinRwinSent: 5840 NagleEnabled: 1 OtherReductions: 0 PktsIn: 15426 PktsOut: 30845 PktsRetrans: 2 RcvWinScale: 7 SACKEnabled: 3 SACKsRcvd: 185 SendStall: 0 SlowStart: 268 SampleRTT: 51 SmoothedRTT: 51 SndWinScale: 2 SndLimTimeRwin: 2050808 SndLimTimeCwnd: 7950685 SndLimTimeSender: 52122 SndLimTransRwin: 4 SndLimTransCwnd: 15 SndLimTransSender: 11 SndLimBytesRwin: 9481152 SndLimBytesCwnd: 35874112 SndLimBytesSender: 46336 SubsequentTimeouts: 0 SumRTT: 684402 Timeouts: 0 TimestampsEnabled: 0 WinScaleRcvd: 2 WinScaleSent: 7 DupAcksOut: 0 StartTimeUsec: 777958 Duration: 10053839 c2sData: 3 c2sAck: 3 s2cData: 4 s2cAck: 5 half_duplex: 0 link: 100 congestion: 0 bad_cable: 0 mismatch: 0 spd: 36.13 bw: 43.33 loss: 0.000032420 avgrtt: 44.91 waitsec: 0.00 timesec: 10.00 order: 0.0119 rwintime: 0.2040 sendtime: 0.0052 cwndtime: 0.7908 rwin: 1.9940 swin: 5.1591 cwin: 2.0051 rttsec: 0.044905 Sndbuf: 676216 aspd: 0.00000 CWND-Limited: 344.03 minCWNDpeak: 2904 maxCWNDpeak: 262812 CWNDpeaks: 1 The theoretical network limit is 43.33 Mbps The NDT server has a 330.0 KByte buffer which limits the throughput to 114.88 Mbps Your PC/Workstation has a 255.0 KByte buffer which limits the throughput to 44.40 Mbps The network based flow control limits the throughput to 44.65 Mbps Client Data reports link is 'Ethernet', Client Acks report link is 'Ethernet' Server Data reports link is 'T3', Server Acks report link is 'FastE'

[SecTSys]

Aye very similar results to what i was getting, - you need an engineer to come out and test and check those faults i am afraid. -

GL with that!

[Ronski]

Thanks for your reply, just about to raise a fault with Plus Net and have noticed that the BT Wholesale checker has now lowered it's estimate to:

Our test also indicates that your line currently supports a fibre technology with an estimated WBC FTTC Broadband where consumers have received downstream line speed of 47.8 Mbps and upstream line speed of 10.9 Mbps.

Now this was previously 57/20 so they must be updating the speed estimate database with real world data from my connection.

[SecTSys]

lmao - they will try to cover their buts with the line profile correction features that supposedly drop when they detect your line is unstable, but what that generally means is there is a fault on the line and it is up to you to ask us to come out and deal with it.  and to be honest this behavior makes me sick. If the profile is dropping because it is unstable there should be a warning at about 25 meg not at 10 meg as their currently is... but that is my opinion.

From the seems of it you need to get a lift and shift on the DSLAM because one of the connections is faulty, and have your line profile reset - this should get you back to the full 80/20 that is available on the line though realistically you should be looking at a 73/15 on the speed tests.  unless you have a slightly slower package that is!

If Plus net ask for proof - send them the network diagnostic data, though i doubt they will ask. - they will run their own tests and swear blind that there isn't a problem, in which case you need to throw the book at them... - (In my opinion it's worse dealing with the English "Yorkshire" call center than it is BT's Indian one!)

3.242E-5 packets lost during test
185 selective acknowledgement packets received

this is awful!!! and there will always be a packet test of 250 packets where is the rest!

[Ronski]

Thanks, I've added the network diagnostic data to my thread on Plus Net.

Strangly the checkers are showing the 57/20 estimate this morning, how strange it was different last night, but I am checking from work, shoudn't make any difference but one does wonder.

[Ronski]

Plus Net at first said it was within limits, but then Alex checked and said it wasn't right.

So I have an engineer coming out Thursday PM, lets hope he knows his stuff/is not on a tight schedule/and can do a lift and shift if required.

[SecTSys]

Indeed - glad that the line stats have checked out for you and the engineer is on his way.

 

[Ronski]

The strange thing is that the modem hasn't resynced since last Thursday, which is now the longest period it's stayed up for, previous longest was about 3 days, with all the others being far shorter.

Guess we'll just have to wait and see if he find's anything.

[SecTSys]

yup - a few things -
1: get the faceplate checked. (sometimes the simple things are missed)
2: get the copper wire pairing checked.
3: get the connections between each connection and the connections themselves checked (on my line there are 5 between my faceplate and the fiber cabinet 6 i believe to the exchange)
4: on a fiber connection there are 2 cabinets that your connection passes through - the copper line cabinet which is then connected to the fiber cabinet. (i believe this then loops back to the copper cabinet but don't quote me on that!)

Whilst this is all happening - when the engineer pauses for a quick break after running tests run the measurement labs tests - if the internet is connected (sometimes this can help the engineer out a little if he looks bewildered!)

4: before he leaves run speed tests at measurement labs. - just to check that everything is working alright. - and if there is jitter or other problems detected then whilst the engineer is there he can look into that too.  - I think this is why i like Russ so much as an Engineer. he appreciates the quality of his work.

[Ronski]

Thanks for the pointers - very useful.

Something odd just happened to my connection, was looking at my ping graph from work and noticed a resync arround 11:00am with an increase in ping - interleaving must have been increased, but the speeds have also altered, down has increased while up has decreased.

So I download my log file from my server, loaded it into a databas and found the following:

Code: [Select]

Date Time Sync Down Sync Up Attainable Down Attainable Up

18/09/2012 11:00 42684 10088 50908 11539
18/09/2012 11:01 42684 10088 50908 11305
18/09/2012 11:02 42684 10088 50908 11522
18/09/2012 11:03 42684 10088 51032 7413
18/09/2012 11:04 42684 10088 51156 7486
18/09/2012 11:05 49237 6445 57880 7948
18/09/2012 11:06 49237 6445 58492 6554
18/09/2012 11:07 49237 6445 58492 6537
18/09/2012 11:08 49237 6445 58492 6495
18/09/2012 11:09 49237 6445 60780 6629
18/09/2012 11:10 49237 6445 60720 8616
18/09/2012 11:11 49237 6445 60124 7258
18/09/2012 11:12 49237 6445 60104 7221
18/09/2012 11:13 49237 6445 59180 7355
18/09/2012 11:14 49237 6445 59196 7444
18/09/2012 11:15 49237 6445 59080 7379
18/09/2012 11:16 49237 6445 59088 7406
18/09/2012 11:17 49237 6445 59088 7327
18/09/2012 11:18 49237 6445 60884 7441
18/09/2012 11:19 49237 6445 60844 7458
18/09/2012 11:20 49237 6445 60936 7334
18/09/2012 11:21 49237 6445 60904 7337
18/09/2012 11:22 49237 6445 60880 7344
18/09/2012 11:23 49237 6445 60856 7355
18/09/2012 11:24 49237 6445 60720 7310
18/09/2012 11:25 49237 6445 60708 7382
18/09/2012 11:26 49237 6445 60696 7379
18/09/2012 11:27 49237 6445 60684 7389
18/09/2012 11:28 49237 6445 60788 7313
18/09/2012 11:29 49237 6445 60668 7313
18/09/2012 11:30 49237 6445 60776 7310
18/09/2012 11:31 49237 6445 59036 7221
18/09/2012 11:32 49237 6445 58120 7118
18/09/2012 11:33 49237 6445 58128 7076
18/09/2012 11:34 49237 6445 58252 7128
18/09/2012 11:35 49237 6445 58380 7128
18/09/2012 11:36 49237 6445 58380 7052
18/09/2012 11:37 49237 6445 58128 7052
18/09/2012 11:38 49237 6445 58252 7038
18/09/2012 11:39 49237 6445 58128 7063
18/09/2012 11:40 49237 6445 58252 7049
18/09/2012 11:41 49237 6445 58128 7032
18/09/2012 11:42 49237 6445 58128 7011
18/09/2012 11:43 49237 6445 58004 7025
18/09/2012 11:44 49237 6445 57880 7028
18/09/2012 11:45 49237 6445 58128 7011
18/09/2012 11:46 49237 6445 58380 7018
18/09/2012 11:47 49237 6445 58380 7008
18/09/2012 11:48 49237 6445 58256 7001
18/09/2012 11:49 49237 6445 58504 7032

Now looking at the log file in the database I can see what the higest and lowest figures are very easily:

Fastest attainable rate Down: 60936 today at 11:20
Fastest attainable rate Up:17672 on 29th August at 09:06

Fastest sync rate Down: 54533 04:55 to 21:09 on 23 August
Lowest sync rate Down: 42684 6:03 13 September  to 11:05 today

Fastest sync rate Up: 14377 09:06 29 August to 05:53 2 September
Lowest Sync rate Up: 6445 Today at 11:05

Suppose I'd better do some work 

[SecTSys]

aye - i guess you had better. 

[Ronski]

The engineers just been, carried out quite a lot of tests - was here for about 90 minutes. He found no problems, he was going to carry out a lift and shift (change the pairs at the cab) but he was told not to by someone he phoned. They are going to reset the profile, but I don't think this will make any difference to the constant re-syncs (time will tell of course) or the strange episodes.

He didn't want to look at any graphs as he said it all went over his head 

[Ronski]

Just had a phone call from the engineer, on the way out the estate he passed the PCP/DSLAM and someone was working on it, so he stopped. Turns out it was a technical officer, remaking connections as there had been a lot of problems with that particular cabinet. So hopefully that will solve the problem, may also explain why the engineer didn't see any error faults in his tests. He did spot the modem re-sync (DSL light went off) when he first arrived, and the guy at the cabinet said that was about the time when he started working on it.

Edit: Didn't manage to find out  the line length - he wasn't running that test, but he did say they we're copper cables :-)