Why I regret joining virgin media

[Sephiroth]

Originally Posted by Ignitionnet

Seph is great for putting things into focus and pointing out when things aren't as they should be .......Google would tell you the answer in a few minutes as those 9 digit numbers do stick out a touch in the modem stats.

A quick Google for "Correctable Codewords" brought up http://www.dslreports.com/forum/r237...internet-speed amongst other matches.

....I tend to leave you in Seph's capable hands to look after you as only he can

Leaving your (magnanimous or faint?) praise in for my own satisfaction - I took a look at the DSLReports thread to see what I would have said to that OP.

I would have noted that the uncorrectables were only 4/100 % (.0004) of the total codewords and would not have considered that worth mentioning other than to say just that.

I would have picked up on the 32 dB SNR for 256QAM and asked to see the event log in order to see what was going on for correlation purposes.

At that stage I would have had little doubt that SNR was stunting the download speed. I would have been puzzled at the low correctabhles count because my knowledge of what goes on further up the network in terms of downstream error correction is nil. Might have said as much - dunno.

[Ignitionnet]

It was by a couple of posters - http://www.dslreports.com/forum/r23786986- // http://www.dslreports.com/forum/r23787134-

Unsure what it ended up being, worth noting that he has a DOCSIS 3 modem but is not provisioned to bond channels, and that they are using US DOCSIS not EuroDOCSIS so more tolerance to noise, 6MHz wide channel instead of 8, and 32dB is borderline instead of being plain rubbish.

[speedfreak]

Love the banter between you two very entertaining and I get to learn at the same time!


Right, Ive realised something. Having an ongoing fault for this long has possibly made me over analyse everything with the 20Mb. Now that Im on 20Mb I think its fine. Running pingplotter 24/7 waiting for spikes just isnt the way to go. Im sure it wouldnt be any different no matter what ISP I'm with, its bound to spike at some points and to be honest at least I dont suffer with peak time congestion like I did with AOL and TalkTalk. For now I will stay with 20Mb until the 50 is fixed, aslong as I can game and not be traffic managed Im happy.

The only worrying thing is that my faults on the 50 service dont tie in with the moto issue, I dont have stacks of T3's and after the faults they did find, no T4's or dropped connections. The codeword issue doesnt seem to be a reported thing with the moto issue so I could wait however long for the moto issue to be fixed, rejoin the 50Mb and find Ive still got a fault. They have ran out of things to look for that is causing my issues so they can only think its motorola related. I guess time will tell.

On the 20Mb I'm told I get around 15000 codeword errors per day (cant check myself on this cisco modem) which they said is nothing to worry about. For now instead of looking for faults Im going to see how it is if I just use the service like any normal person would Fingers crossed the 50Mb gets sorted soon though, it 100% isn't right, no over analysis went on there! Gaming was awful at times.




P.S..... Pabs, re: the avatar, anti wrinkle cream is my secret to keeping myself looking so young, skin like a 2yr old for just £9.99 per tub 4 days off work now so I will be hammering BFBC2

[Sephiroth]

Originally Posted by Ignitionnet

There's some downstream RF issues that don't show up on the SNR reported by the modem, same goes for upstream in the other direction too.

Challenge is to find someone at VM who'll actually persue the matter for you.

I think we'll be back to this. You gonna show us your stats etc again now you're on 20 meg?

[speedfreak]

This modem doesnt show as much, I posted them earlier but I'll post again incase I missed anything out, Im told Ive only had around 8 T3's in nearly 2 days so no problem there and my power levels are bang on, like I say Im just worried I could wait months for a motorola fix only to find I still have a 50Mb fault

Had this but no powercycle of the CM needed so I guess nothing to worry about ...
Time Level Description
Wed Apr 28 13:49:40 2010 Critical (3) Started Unicast Maintenance Ranging - No Response received - .

attached a JPEG so its abit easier to make out than random text everywhere

I guess with VM exhausting all avenues and you guys not having access to VM's logs theres not alot of advice you can give.

[speedfreak]

ok I really should wait longer before posting after a few hours gaming this connection is even worse I really think I have other issues not related to the motorola fault. The pain now is I've got 4 days off work, plenty of time for gaming but the CEO guy is off till tuesday so I'm stuck with noone to switch me back arghhhhhh! No idea where I go from here but for now I NEED the 50 back!

[Sephiroth]

You don't need 50 meg for gaming. 1½ meg downstream is perfectly good (as we've proved with our ADSL connexion).

What you do need is reliable on demand upstream timeslots. The 50 meg users share 1 single 9 meg upstream channel per node (Igni might say 1 or 2 nodes). That's absolutely bugger all when people get busy gaming and P2P-ing. Gaming needs frequent timeslots of small data packets - but very frequent and you don't want to be waiting for everyone else.

Actually I think that if you're on the newer overlay network the 20 meg users have the same upstream problem. Maybe Igni can confirm or otherwise.

[speedfreak]

Originally Posted by Sephiroth

You don't need 50 meg for gaming. 1½ meg downstream is perfectly good (as we've proved with our ADSL connexion).

What you do need is reliable on demand upstream timeslots. The 50 meg users share 1 single 9 meg upstream channel per node (Igni might say 1 or 2 nodes). That's absolutely bugger all when people get busy gaming and P2P-ing. Gaming needs frequent timeslots of small data packets - but very frequent and you don't want to be waiting for everyone else.

Actually I think that if you're on the newer overlay network the 20 meg users have the same upstream problem. Maybe Igni can confirm or otherwise.

hey I'm not that simples I know I dont need 50 down for gaming but this connection is even worse. I just dont understand why Your post probably explains it but its lost on me. If you ever have a boiler issue ask me for help and I wont baffle you with science If I can be pointed in the direction of potential problems/solutions I'm certain CEO guy would look in to it, its a problem when networks tell him theres no issue though

[Ignitionnet]

Originally Posted by Sephiroth

You don't need 50 meg for gaming. 1½ meg downstream is perfectly good (as we've proved with our ADSL connexion).

What you do need is reliable on demand upstream timeslots. The 50 meg users share 1 single 9 meg upstream channel per node (Igni might say 1 or 2 nodes). That's absolutely bugger all when people get busy gaming and P2P-ing. Gaming needs frequent timeslots of small data packets - but very frequent and you don't want to be waiting for everyone else.

Actually I think that if you're on the newer overlay network the 20 meg users have the same upstream problem. Maybe Igni can confirm or otherwise.

Yes sir - anyone on the overlay network, regardless of tier, uses the same upstream capacity.

If speedfreak could paste his downstream frequency it would confirm which network he's connected to.

[speedfreak]

Originally Posted by Ignitionnet

Yes sir - anyone on the overlay network, regardless of tier, uses the same upstream capacity.

If speedfreak could paste his downstream frequency it would confirm which network he's connected to.

is that what image is attached to post 110? soz on the ps3 and its hard to look things up

[pabscars]

Originally Posted by speedfreak

hey I'm not that simples I know I dont need 50 down for gaming but this connection is even worse. I just dont understand why Your post probably explains it but its lost on me. If you ever have a boiler issue ask me for help and I wont baffle you with science If I can be pointed in the direction of potential problems/solutions I'm certain CEO guy would look in to it, its a problem when networks tell him theres no issue though

Sorry to hear your still knee deep in shmit speedy, and just for comparison, I,m on the 20Mb tier for a bit as you know, because they wanted to get me off upstream port 1 (docsis 3) due to it being mullered.

I,m now on docsis 1 upstream port 3 which at worst only reaches 30% capacity, but unfortunately the downstream is pretty hammered between 7 till 11pm, which may or may not explain the continued spikes.

After Igni posted a reply re my jitter, Ive been delving more into it, and I think this is more likely to be the cause of the problems I,m experiencing, maybe its the same with you.

I dont know if its true but I presume the more hammer your line gets, the more jitter is produced, and therefore unpleasant gaming (bad analogy I know but it works for me).


I noticed yesterday on MW2 that at 6pm I had a quick game and apart from a few minor spikes all was good, it was much easier to shoot opponents, and I finished top of the board with 26 kills to 5 deaths.

Switched back on about 9 ish and it was crap again so I was a bit peeved to say the least. But trying not to be overly objective I played all night to give it some stick, and by 11:30 I was ready to smash the lot to bits so switched it off.

Gaming at times is a little better but still get the same old spikes and its worse in the evenings, I just feel like you probably; that theres no more avenues to go down, nobody left to ask, and you get sick of complaining after a bit.

The only difference between me and you is location, we both get the same issues on the same games at the same time. So one can only really draw one conclusion from that IMO, and that is that the vm network in parts is so overstretched/contended that is causes unstable pings, loads of jitter at peak times which plays havoc on FPS games.

Re your 50Mb modem, are you sure its not still provisioned, because apparently mine is so if it gets that bad I can try switching out the modems and it should work, therefore you might be in the same position.

Give it a wurl it might work.

I,m gutted for you pal, because I know just how bloody frustrating it all is.

Lets keep hoping for some sort of miracle (ever the optimist) ,

Have you tried any tests on the myvoip site to see how it looks compared to the 50Mb tier?

---------- Post added at 00:43 ---------- Previous post was at 00:41 ----------

Oh almost forgot to mention, where can I get my hands on some of that cream

[Sephiroth]

Originally Posted by speedfreak

hey I'm not that simples I know I dont need 50 down for gaming but this connection is even worse. I just dont understand why Your post probably explains it but its lost on me. If you ever have a boiler issue ask me for help and I wont baffle you with science If I can be pointed in the direction of potential problems/solutions I'm certain CEO guy would look in to it, its a problem when networks tell him theres no issue though

LOL. Trouble is that this HFC stuff is science. It's amazing it works at all!

The upstream channel (you've and the others on your node have got access to 1 upstream channel) is organised into timeslots for which your modem can bid when it needs to send something upstream. When you gain a timeslot, your data up to codeword size goes into the opportunity given by the timeslot. For gaming, the amount of data per request is very small, but the requests are very frequent as you move your mouse or shoot the peasant.

If your area/node is oversubscribed or heavily utilised, your timeslot opportunity doesn't come up as frequently as you need it, so there is lag and jumping around etc.

The solution, as Igni points out, lies in resolving contention at the node end. He said (and this is what the CEO needs to take on board):

Could take another channel from the legacy network and either bond them or run them as a single 6.4MHz channel. Could also clean up network some more and/or use better upstream PHY migrating everything to DOCSIS 2 and use previously unusable frequencies for SCDMA / 16QAM TDMA to add additional capacity.

Reducing node sizes or upstream service groups further will have the effect of cleaning up networks, at present a number of overlay upstreams span more than one node.

These would produce considerably more capacity in the same RF space.

HTH.

[Ignitionnet]

Close though requests aren't made up to 'codeword size' but in mini-slots, and one of the parameters in the upstream MAP is mini-slot size in 'ticks' of 6.25us. The modem asks for as many mini-slots as it needs then receives either a grant for those or a partial grant for some of them with the rest fulfilled when the CMTS can, which happens under congestion conditions. The amount of mini-slots a modem can request at once is set by the maximum upstream burst parameter, which you can probably read off your modem's diagnostic pages. They're allowed enough mini-slots from a single request to transmit that amount of data. Note that modems can, and do, encapsulate further requests for grants along with their data when transmitting across multiple bursts to avoid going through the contended mini-slots again to make their requests.

Concatenation, which you'll see in DOCSIS 1.1 and above and some 1.0 hardware, is also useful, it allows multiple small frames to be placed into a single burst request which seriously increases the packets per second as the modem doesn't have to go through the request-grant-transmit cycle for each and every datagram. Very relevant for very high frequency but small frame size traffic such as gaming.

[Chrysalis]

interesting info on the timeslots, so it is possible I guess to get jitter even when the bandwidth itself isnt saturated but if too many people want a timeslot at the same time.

I too learn from seph.

[Ignitionnet]

Originally Posted by Chrysalis

interesting info on the timeslots, so it is possible I guess to get jitter even when the bandwidth itself isnt saturated but if too many people want a timeslot at the same time.

I too learn from seph.

Unsure what you mean there - could you elaborate? If there are too many outstanding requests for mini-slots that is the definition of saturation on an upstream. Utilisation on upstreams isn't measured in bandwidth terms it's actually measured in % of mini-slots used.

Modem gets data from customer's equipment, requests the appropriate amount of mini-slots. There is the potential for minimal jitter, on the order of a millisecond or two, where the modem sends a request and it collides with a request from another modem on the way up to the CMTS but this is not a significant source of jitter. The major cause is where the CMTS has no mini-slots available for a while, or indeed at all, in this case it'll either give the modem a zero-length grant and the modem will have to re-request the bandwidth again or it'll just not give any response at all.

Where a modem has a 100ms delay it has gone through repeated cycles of requesting bandwidth and receiving zero length grants or none at all until eventually, nearly 100ms later, it finally receives a grant and transmits its' data.

Where a modem shows packet loss as well a buffer in the modem filled up and overflowed, so the data in there for the longest period was dropped to make room for newer data. If congestion is heavy enough it's quite possible for this to happen as the modem will only hold so much data while waiting for upstream grant.