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

[craigj2k12]

Originally Posted by Kushan

It's never ONE guy. One person does not have the ability to congest the entire network.

one user with 10 or 12 meg upload could easily kill 1 upstream channel

[Kushan]

Originally Posted by craigj2k12

one user with 10 or 12 meg upload could easily kill 1 upstream channel

How do you work that one out? Each upstream has about 30mbit of bandwidth. Sure, one person on 10meg upload could saturate over 1/3 of it, but hardly kill it.

[craigj2k12]

Originally Posted by Kushan

How do you work that one out? Each upstream has about 30mbit of bandwidth. Sure, one person on 10meg upload could saturate over 1/3 of it, but hardly kill it.

20

[Kushan]

Originally Posted by craigj2k12

20

30

http://en.wikipedia.org/wiki/DOCSIS#...onal_standards

n × 30.72 (n × 27) Mbit/s

[craigj2k12]

Data rate = #_Bits_per_symbol * Symbol_Rate

Bits per symbol = 4
Symbol Rate = 5120000

= 20,480,000

or 20.48Mb/s

---------- Post added at 13:30 ---------- Previous post was at 13:30 ----------

http://volpefirm.com/blog/docsis-101...1_upstream-rf/

[Kushan]

Originally Posted by craigj2k12

Data rate = #_Bits_per_symbol * Symbol_Rate

Bits per symbol = 4
Symbol Rate = 5120000

= 20,480,000

or 20.48Mb/s

---------- Post added at 13:30 ---------- Previous post was at 13:30 ----------

http://volpefirm.com/blog/docsis-101...1_upstream-rf/

So it's because Virgin uses QAM16, then? Today I learned.

[Eeeps]

Originally Posted by Kushan

So it's because Virgin uses QAM16, then? Today I learned.

I guess it's the poor state of the field plant and the number of modems per channel that doesn't let them get to QAM64.

I see a lot of cut cables outside old VM customer houses in my area. I bet they haven't been disconnected at the cab.

My upstream power is all over the place from 41 to 44dBmV

[Mick Fisher]

Originally Posted by Qtx

Virgin have won when they get customers blaming others for their poor connection. Virgin not supplying enough up or downstream for the amount of customers connected is where I would lay the blame.

Exactly!!

[qasdfdsaq]

Originally Posted by Sephiroth

The takeaway in the context of Qasi's post is that the frequency of ACKs sent back (through a busy) upstream on data received will slow the download (decrease throughput).

Whether or not that is a 5-10% overhead I'll leave to Qasi to justify, if he cares to. But it is an important point that he makes.

Whoops, missed this earlier (haven't been on much lately)

TCP overhead is typically 1.5-3% of the downstream usage. Then add on top DOCSIS overheads, and the fact an upstream channel has really about 1/3rd the capacity of a downstream channel, that means 4.5-9% of the upstream channel I just rounded that to 5-10%.

---------- Post added at 17:22 ---------- Previous post was at 17:16 ----------

Originally Posted by Kushan

Believe it or not, this situation is nothing new to VM. Remember when the 50Meg first came out and had an upload speed of barely over 1.5Mbit? (If I recall correctly). The ACKs alone would nearly congest the upload stream in itself, yet it actually ran fairly smoothly (For most, there's always an exception).

Heh, well... On 50Meg they had 2 upstreams for every area (theoretically), but at half the bandwidth, so one upstream per 4 channels. What they have now, I don't know, but if it is 1-between-8, then that is not a good thing given general utilization in both directions will be higher now than it was when 50 came out.

50 didn't run particularly smoothly in areas with high utilization - engineers in my area quoted 90% capacity in use 90% of the time (!) - and getting more than 300Kbps upload was a chore some days.

Anyway, there is a point in what you're saying but he's having this issue without actually utilising his downstream, which is the point I'm making - just HAVING 8 downstreams versus 1 upstream isn't the issue, there's obviously something else going on there.

The problem isn't him using his connection, both upstream and downstream channels are shared between hundreds of users. Other people using it to capacity will cause problems for him whether he uses it himself or not.

I'd like to know where you're getting your figures, though, 5-10% of upstream per downstream for example. Are you referring to the available bandwidth for the connection (What profile the modem is set to), or the available bandwidth for the channels (8x50Mbit on the downstream, 1x~30Mbit on the upstream?) or what?

Bandwidth over the whole channel (8x55Mbit on the downstream, 1x20Mbit on the upstream).

Remember upstream is inherently inefficient on cable, because of the requirement for CSMA and guard intervals, etc. and you will rarely be able to use entirety of a channel's upstream capacity. The practical maximum is closer to 16-18Mbit available for actual use.

55Mbps dowstream with 2% used for ACKs is around 1.1Mbps if everything behaves efficiently, depending on RWIN, MTU, OS, etc. this can vary both up and down. Don't forget also a single 100/10 user could eat up over half the upstream capacity on such a channel.

(An ACK takes around 54 bytes 'on the wire', which with a pessimistic minimum MTU of 576 bytes, and no ack suppresion, etc., means 8 downstream channels could actually require ~40Mbps of upstream capacity in an almost-worst-case-scenario, without anybody doing any actual data uploads)

[craigj2k12]

Originally Posted by qasdfdsaq

55Mbps dowstream with 2% used for ACKs is around 1.1Mbps if everything behaves efficiently, depending on RWIN, MTU, OS, etc. this can vary both up and down.

So on a channels with (in best case) 18Mbit useable bandwidth, on 8 downstreams, 8.8Mbit could be used just on acknowledgements, then there are users expecting to get 10 or 12Mbit upload speeds

[qasdfdsaq]

Originally Posted by roughbeast

LOL. Not even I, am that nerdy.

The obsessive speedtester need not necessarily be you - remember, all capacity is shared. Some other person obsessively speedtesting (and using up over 50% of the network capacity at such a time) could well be the cause.

---------- Post added at 17:35 ---------- Previous post was at 17:34 ----------

Originally Posted by Kushan

It's funny how the worse the service gets, the prettier the graph is...

Derp:


---------- Post added at 17:37 ---------- Previous post was at 17:35 ----------

Originally Posted by craigj2k12

So on a channels with (in best case) 18Mbit useable bandwidth, on 8 downstreams, 8.8Mbit could be used just on acknowledgements, then there are users expecting to get 10 or 12Mbit upload speeds

Best-case involves ACK suppression (VM don't bother with this) SACK's, zero packet loss and large RWINs that reduce ACK overhead to well under 1%.

The 1.5% - 3.0% overhead I quote is for typical real-world scenarios.

[Edit]
Never mind, you meant at best 18Mbps useable upload bandwidth, yes, right.

That said I'm only speculating on the number of upstream channels. Some people supposedly report their area only has one, but in the past common practice was to have at least two (at that time they were half the capacity per channel though)


---------- Post added at 17:40 ---------- Previous post was at 17:37 ----------

Originally Posted by Kushan

It's never ONE guy. One person does not have the ability to congest the entire network.

Yes they do.

A "one guy" with 10Mbps upload speed on a 20Mbps channel can easily cause congestion - congestion can become visible at as low as 60-70% utilization on the upstream. All other users need only be using 30% of available capacity between them before one 100/10 user can "congest the entire network"

---------- Post added at 17:41 ---------- Previous post was at 17:40 ----------

Originally Posted by craigj2k12

I bet the techs have projectors flicking through slides of pretty graphs at hubsites.

Heres what the Bristol hubsite, looks like

From the outside, it is merely a standard looking hubsite...



...but inside......

Way to advertise... "Press here to cripple our network".

---------- Post added at 17:42 ---------- Previous post was at 17:41 ----------

Originally Posted by Kushan

So it's because Virgin uses QAM16, then? Today I learned.

Up until a few months ago some areas still used BPSK. The BPSK => 16QAM upgrade only started in 2011.

That said till recently the Superhub also clearly displayed 20Mbps next to the upstream channels on its normal status pages.

[Sephiroth]

Originally Posted by qasdfdsaq

......Best-case involves ACK suppression (VM don't bother with this) SACK's, zero packet loss and large RWINs that reduce ACK overhead to well under 1%.

The 1.5% - 3.0% overhead I quote is for typical real-world scenarios.
.......

AFAIK, VM used TAS (Ack Suppression) in the form of Turbodox. I don't know the poosition now, but I imagine it could still be in use. Here is a supporting quote from Mark Wilkin in September 2010:

We've just released the new 12.01 firmware update for VM300 Cable Modems across our network.

The new firmware was tested over the past 8 weeks with 2000 customers including volunteers from this forum and we've seen no significant issues in testing.

The new firmware enables 100Mbps throughput for our forthcoming 100Mbps service and is also designed to improve upstream stability and throughput. It also turns back on the TDOX feature that was switched off in July with the 10.12 firmware, due to a bug affecting performance.

BTW, note VM saying that the VMNG300 was fit for 100 meg.

[Chrysalis]

Originally Posted by Kushan

It's never ONE guy. One person does not have the ability to congest the entire network.

it could be if eg. the normal utilisation is say 60% and then a top tier user with 10 or 12 upload goes full throttle. On a 18mbit US channel one user can use more than half of the bandwidth, a 12mbit customer can use 2/3 of a US channel by them self. Your comment is right if the normal utilisation is 0% which isnt the case. I have been surprised to hear reports VM have been activating 120/12 in areas with no US bonding, seems nuts.

---------- Post added at 18:12 ---------- Previous post was at 18:09 ----------

Originally Posted by Qtx

Virgin have won when they get customers blaming others for their poor connection. Virgin not supplying enough up or downstream for the amount of customers connected is where I would lay the blame.

Ultimately thats the problem, in particular too many modems per service group, but also that isnt fat enough pipes at the docsis level as well. Any setup that has a customer able to use more than half of a shared pipe is going to be fragile even at 2:1 contention.

[qasdfdsaq]

Originally Posted by Chrysalis

it could be if eg. the normal utilisation is say 60% and then a top tier user with 10 or 12 upload goes full throttle. On a 18mbit US channel one user can use more than half of the bandwidth, a 12mbit customer can use 2/3 of a US channel by them self. Your comment is right if the normal utilisation is 0% which isnt the case. I have been surprised to hear reports VM have been activating 120/12 in areas with no US bonding, seems nuts.

As far as I know 12Mbit upload is never implemented before 2-channel bonded upstream is implemented. In other words, 40Mbps upload channels are a prerequisite before anybody gets 12Mb

[Chrysalis]

Originally Posted by Sephiroth

AFAIK, VM used TAS (Ack Suppression) in the form of Turbodox. I don't know the poosition now, but I imagine it could still be in use. Here is a supporting quote from Mark Wilkin in September 2010:

We've just released the new 12.01 firmware update for VM300 Cable Modems across our network.

The new firmware was tested over the past 8 weeks with 2000 customers including volunteers from this forum and we've seen no significant issues in testing.

The new firmware enables 100Mbps throughput for our forthcoming 100Mbps service and is also designed to improve upstream stability and throughput. It also turns back on the TDOX feature that was switched off in July with the 10.12 firmware, due to a bug affecting performance.

BTW, note VM saying that the VMNG300 was fit for 100 meg.

Qas are you sure large rwins reduce ack overhead? The only thing I have found that affects ack overhead is delayed acks, nagle algorithm and adjusting mtu(mss) size. Whilst boosting rwin allows higher throughput the ack throughput goes up with it. Also supressing acks isnt without consequence, its good for certian bulk downloads but bad for small packets. sacks also sends the same amount of acks, the difference sacks offers is just how recovery works if packets are dropped.

---------- Post added at 18:19 ---------- Previous post was at 18:19 ----------

Originally Posted by qasdfdsaq

As far as I know 12Mbit upload is never implemented before 2-channel bonded upstream is implemented. In other words, 40Mbps upload channels are a prerequisite before anybody gets 12Mb

Thats what I thought until my neighbour was put on 12mbit without bonding and I seen reports on tbb with the same.