50M : Think Broadband Ping Monitor Results (POST YOURS)

[Kymmy]

Originally Posted by qasdfdsaq

Given that O2/Be backbones and certain lines to France were also affected, this incident does seem to have a bit of a wider scope than "very close to each other".

So why would a ping from only 80.249.99.164 which is only 3 jumps outside of VM's network be effected by a download of iOS6? Is there any history for a software release to effect ping times across a couple of countries?

Surely (yes I am calling you Shirley ) the server bottlenecks would have stopped the backbones getting saturated unless of course the requests themselves are the cause? Perhaps you can explain more..please..

[craigj2k12]

Does anyone know where the bottleneck was, i.e. did anyone so a few tracerts last night. Maybe people across the world downloading iOS6 saturated a couple of peering links?

[Jayster]

Originally Posted by ferretuk

I assumed the reference to be the impact of mass downloading of IOS 6 globally on Tbb ping monitoring - on topic?

I was, sorry if it was a bit vague.

[qasdfdsaq]

Originally Posted by Kymmy

So why would a ping from only 80.249.99.164 which is only 3 jumps outside of VM's network be effected by a download of iOS6? Is there any history for a software release to effect ping times across a couple of countries?

Surely (yes I am calling you Shirley ) the server bottlenecks would have stopped the backbones getting saturated unless of course the requests themselves are the cause? Perhaps you can explain more..please..

I didn't say anything about iOS6, I only said the issue was more widespread than "very close". As for jumps - the most critical jumps don't show up on traceroutes. And lest not forget, Moscow and New York are "only" 3 jumps outside of VM's network as well.

While I little care about iOS6 or its effects on the internet, I think you've got server bottlenecks and backbones the wrong way round. Servers have far more capacity than the switches they are connected to, and those switches have far more capacity than the routers or backbones they are in turn connected to.

For example one datacentre may have 1000 servers with 1-4Gbps capacity each, hooked up to 64 10Gbps switches hooked up to 40Gbps of internet.

[scaseman]

http://www.thinkbroadband.com/ping/s...0-09-2012.html

Looks like it's doing it again tonight.

[ileikcaek]

This is getting ridiculous... One night, sure I can live with that, but two in a row is just plain unacceptable. my speeds are decreasing as the ping on that chart gets higher. 55ms right now with a speed of 90 down 8.87 up...

[img]Download Failed (1)[/img]

[Kymmy]

Same here again so that discounts iOS6 downloads but points towards something being altered/started at the exact same time

---------- Post added at 19:11 ---------- Previous post was at 19:10 ----------

Originally Posted by qasdfdsaq

I didn't say anything about iOS6, I only said the issue was more widespread than "very close". As for jumps - the most critical jumps don't show up on traceroutes. And lest not forget, Moscow and New York are "only" 3 jumps outside of VM's network as well.

While I little care about iOS6 or its effects on the internet, I think you've got server bottlenecks and backbones the wrong way round. Servers have far more capacity than the switches they are connected to, and those switches have far more capacity than the routers or backbones they are in turn connected to.

For example one datacentre may have 1000 servers with 1-4Gbps capacity each, hooked up to 64 10Gbps switches hooked up to 40Gbps of internet.

Thanks for the more detailed explanation

[ileikcaek]

looks to me like the graphs are normalizing already so I am guessing VM must have changed the routing around what ever is causing the issue?

[iateallthepies]

---------- Post added at 20:41 ---------- Previous post was at 20:40 ----------

Originally Posted by iateallthepies

Mine looks very similar to the one above again..

[Blairhoyle]

Originally Posted by rickymallory

This is getting ridiculous... One night, sure I can live with that, but two in a row is just plain unacceptable. my speeds are decreasing as the ping on that chart gets higher. 55ms right now with a speed of 90 down 8.87 up...

I had the same high Minimum Latency last night it seems.

any idea what causes that? been an ongoing issue over the last week

[Chrysalis]

I got the 6-8pm spike but I also got a load of US congestion around it followed by ongoing congestion to now till nearly 2am, so looks like my new US channel didnt work well for long, a torrenter has followed me onto the channel. Seems the spike made others channel hop to try and get round it.

[Martin_D]

Getting a little bit of Packet Loss

http://www.thinkbroadband.com/ping/s...9c82e93f6.html

[RealDiamond]

The routing fix seams to be holding this time. The default gateways even back to normal..Might still be Knacked else were... but normal service no spike yet... for me..

[Neo-Tech]

This is what happens when you don't pay attention to your monitor for a few days!
Gone from 5 downstream to 6 downstream, and from 1 upstream, to 2 upstream. Stats and graphs below.

Downstream Channels
Lock Status Channel ID Frequency Modulation Rx Power SNR Pre RS Errors Post RS Errors
Locked 35 299000000 Hz QAM256 0.4 dBmV 39.4 dB 48026 78
Locked 33 283000000 Hz QAM256 0.7 dBmV 39.1 dB 60526 104
Locked 34 291000000 Hz QAM256 0.8 dBmV 39.4 dB 114783 75
Locked 36 307000000 Hz QAM256 -0.9 dBmV 38.2 dB 516064 24
Locked 37 315000000 Hz QAM256 0.1 dBmV 38.5 dB 125264 30
Locked 38 323000000 Hz QAM256 0.0 dBmV 38.3 dB 27727 9

Upstream Channels
Lock Status Channel ID Frequency Modulation Tx Power Mode Channel Bandwidth Symbol Rate
Locked 43 27400000 Hz ATDMA 35.5 dBmV 16QAM 6400000 20480 Kbits/sec
Locked 42 35800000 Hz ATDMA 37.3 dBmV 16QAM 6400000 20480 Kbits/sec


(No idea what happened here, judging by the comments on this thread, looks like a TBB issue.)



(Virgin Media did well with not making me notice until after a few days. Had no idea the internet had been down, usually I do!)




Haven't really noticed a difference in browsing or anything, though.

[RealDiamond]

@ Neo-Tech
You should be ready for 60/6 soon then.
let us know how it goes... might want to upgrade..