120M : Best way to escalate a utilisation fault?

[qasdfdsaq]

Originally Posted by Kushan

While I'm not going to claim that Virgin don't often take the **** when it comes to dealing with utilisation, lately they've been better at it so I wonder what's really going on in your area. There must be some complication, or some red tape causing delays.

Same is true of all ISPs, mobile providers, even enterprise/business leased lines.

VM have indeed improved a great deal over the past few years, but are still far from perfect. Similarly though, even their best competitor (in terms of lowest congestion) isn't perfect either and every ISP in the country has at some point had some congestion somewhere.

---------- Post added at 13:07 ---------- Previous post was at 13:06 ----------

Originally Posted by ianch99

What do you think would be the cause of the 2 year delay in my case?

Pile of poop network? My (old) area always used to be scheduled last (i.e. 18 months behind everyone else) for upgrades and always ran behind schedule too, both local techs and national planners have described it as due to the network in said area as being... lets say... sub-par.

[Kushan]

I've heard a few different techs and such claim that such and such an area was "crumbling apart".

[ianch99]

Interesting development: email from Access HeadEnd Manager – South Coast and South London via CEO/Faults office:

We have some resegmentation work scheduled for Wednesday 20th August to provide 4 additional downstream channels which will increase the amount of available bandwidth for the customer.

9am today, modem reboot and I have new wan ip because I am now on a new CMTS (sotn13)

Power levels and d/s SNR have changed slightly, possibly for the worse? Time will tell I guess:

Downstream DS-1 DS-2 DS-3 DS-4 DS-5 DS-6 DS-7 DS-8
Frequency (Hz) 298750000 266750000 274750000 282750000 290750000 306750000 314750000 322750000
Lock Status(QAM Lock/FEC Sync/MPEG Lock) Locked Locked Locked Locked Locked Locked Locked Locked
Channel ID 9 5 6 7 8 10 11 12
Modulation 256QAM 256QAM 256QAM 256QAM 256QAM 256QAM 256QAM 256QAM
Symbol Rate (Msym/sec) 6.952000 6.952000 6.952000 6.952000 6.952000 6.952000 6.952000 6.952000
Power Level (dBmV) 4.44 3.61 3.73 4.23 4.46 4.60 4.31 4.23
RxMER (dB) 35.78 34.77 34.77 35.25 35.60 35.97 35.97 35.97
Pre RS Errors 1317 3316 3063 1568 1257 1227 2020 1466
Post RS Errors 286 298 991 309 306 302 942 301

Upstream US-1 US-2 US-3 US-4
Channel Type 2.0 N/A N/A 2.0
Channel ID 50 N/A N/A 51
Frequency (Hz) 39400000 N/A N/A 32600000
Ranging Status Success N/A N/A Success
Modulation 16QAM N/A N/A 16QAM
Symbol Rate (Sym/sec) 5120000 N/A N/A 5120000
Mini-Slot Size 4 N/A N/A 4
Power Level (dBmV) 34.00 N/A N/A 34.50
T1 Timeouts 0 0 0 0
T2 Timeouts 0 0 0 0
T3 Timeouts 8 0 0 0
T4 Timeouts 0 0 0 0

New BQM graph:



with old graph:



New one seems to have a higher base latency Not sure why latency would be up when routing to the same headend ..

[qasdfdsaq]

New graph doesn't look all that great tbh either.

Latency can vary for a whole host of reasons. As you can see from the graph the base latency changes over time as well. Mostly this is due to load balancing within VM's core network, where there are multiple paths from A to B that are not all the same length and traffic is shared between them.

[Ignitionnet]

Not sure that that 12 downstream change has actually worked - you're still on the standard 8 channel frequency plan.

Switch your Superhub off for a few minutes and then reactivate it, see if you get a different set of downstream channels.

[ianch99]

Originally Posted by Ignitionnet

Not sure that that 12 downstream change has actually worked - you're still on the standard 8 channel frequency plan.

Switch your Superhub off for a few minutes and then reactivate it, see if you get a different set of downstream channels.

I am not sure they meant 12 d/s channel for the shub, rather it referred to the number of available downstream channel frequencies?

Have restarted the shub2 a couple of times with no observable change. I think the problem may be in the network beyond the headend. Here are two speed tests, the first is London Namesco. This has always been a reliable fast site for me. The second is a VM site in Brentford but crucially inside the VM network:





I did this test a number of times and the VM site was always 160+Mbps and around 10ms latency. The Namesco site was always variable in speed and never near 150 plus ~3 times the latency

This seems to indicate the CMTS hop is now no longer the gating factor (where is was before). If I can get consistent headline speeds to an internal VM site but not to the internet then this raises question about the network path from the headend to the internet. It may also account for the latency difference.

Here are BQM graphs for two CMTS gateways: the first my former sotn8 gateway and the second my new sotn13 one:





You can see the base latency difference .. and this is not involving the CMTS to cable modem segment. What is strange is that these should be in the same (Sotn) headend so why the latency increase for sotn13? Weird ..

[General Maximus]

Originally Posted by ianch99

I am not sure they meant 12 d/s channel for the shub, rather it referred to the number of available downstream channel frequencies.

I think they have made more channels available in the bonding group so that the cmts can load balance across more channels. As an example. over the last 2 years we have seen bonding groups increase in size from 4, to 5, 6, 7 and 8 channels as VM have been trying to increase capacity and reduce congestion. Even though 8 channels may have been available in my area, when I was on my vmng300 it could only use 4 channels and sometimes they were 1, 2, 3 and 4 and other times it was 1, 2, 5 and 6.
Shub's can only use 8 channels so up and till now shub's have sort of been ahead of the game and been able to fully utilise whatever has been made available to it. As VM haven't provided us with 16 channels modems yet, I think the temporary fix is to make more channels available so instead of locking onto the same 8 congested channels, there are other less congested channels available (as in the case of my vmng300). So now instead of using channels 1-8 all the time, you might use 1, 2, 3, 4, 9, 10, 11, 12 as an example.

[Kushan]

I'm pretty sure my superhub2 has connected to channels 8+ for quite some time now?

[General Maximus]

I didn't mean channels numbers, I was talking about channel quantity. There are channel numbers in the hundreds. I was talking about more than 8 channels (in quantity) being available in a bonding group even through the shub can only lock onto 8 (in quantity) at any one time.

[pip08456]

Originally Posted by General Maximus

I didn't mean channels numbers, I was talking about channel quantity. There are channel numbers in the hundreds. I was talking about more than 8 channels (in quantity) being available in a bonding group even through the shub can only lock onto 8 (in quantity) at any one time.

That surely depends on the quality of the processor in the S/H. Not guaranteed!

[General Maximus]

nope, the cmts is responsible for load balancing and it will allocate the channels to the modem. The shub doesn't decide which channels it wants to use. I hate to say but in this case you are wrong Rom Doll. I still love you though