125M : Vmng300

[zekeisaszekedoes]

Originally Posted by Skie

I think the entire sentence should be the context. Their Superhubs are "perfect" but have higher latency than other devices. So allegedly they are perfect.

This is sorta what I meant, yeah.

Also:

Originally Posted by Nopanic

Just out of interest, why are they "alleged" only when someone claims they work ?

Because there's a good chance that with the proper testing their superhubs would choke in the same way everyone elses does. For example, Masque never did get back to me with the results of a "transfer 5GB+ file from 300Mbps wireless N device to 1Gbps ethernet device" test, which used to break mine all the time. Light, regular usage doesn't always show up the problems it has, perhaps leading to the misinformed conclusion that their superhubs are perfect.

Trust me, even the very best kit I own, hand-picked for the job, isn't perfect and my superhub is very far from top of the heap.

[Stephen]

I have transfered a 6Gb file from my wired mac to my wireless windows lappy this week and it averaged 4.3Mb per sec and did not freeze or crash my superhub. Think some people have just been unlucky with it.

[jb66]

Originally Posted by Stephen

I have transfered a 6Gb file from my wired mac to my wireless windows lappy this week and it averaged 4.3Mb per sec and did not freeze or crash my superhub. Think some people have just been unlucky with it.

Please Do Not Refer To The Hub As Super As It Is Clearly Not Super, Thanks

Mod Edit: Bold is for moderation posts only.

[Stephen]

Originally Posted by jb66

Please Do Not Refer To The Hub As Super As It Is Clearly Not Super, Thanks

Its called the Superhub and for me its been perfectly fine so that is what I shall call it. Thanks

[Stuart]

Originally Posted by jb66

Please Do Not Refer To The Hub As Super As It Is Clearly Not Super, Thanks

It is, however, called the Superhub whether it is super or not. So calling Superhub is correct.

[Peter_]

Originally Posted by jb66

Please Do Not Refer To The Hub As Super As It Is Clearly Not Super, Thanks

Well one is called the hub and the other is called the Superhub and both strangely by the company we both work for, or if you want to call it another name why not use VMDG480 and the will be no need for the moderators to edit posts.

[zekeisaszekedoes]

Originally Posted by Stephen

I have transfered a 6Gb file from my wired mac to my wireless windows lappy this week and it averaged 4.3Mb per sec and did not freeze or crash my superhub. Think some people have just been unlucky with it.

Try transferring it back and forth a few times, if you can be bothered. In my experience eventually it'll freeze up the whole superhub requiring a reboot, even when DHCP is off and it's running as a simple switch.

[qasdfdsaq]

As I've mentioned in another thread, 300mbps wireless 'N' is notoriously difficult to get right, most firmware handles it OK in an ideal/greenfield situation but as soon as there's congestion/noise/interference many will fall over and cripple itself, die, or reboot. Even dd-wrt and Openwrt, considered by some as the best OSF around do this regularly. Many of the products that I know of that work "reliably" in HT + MIMO mode do so by being "bad neighbours" and ignoring the interference-mitigation sections of the wireless spec.

[KenK]

Originally Posted by Masque

Well one is called the hub and the other is called the Superhub and both strangely by the company we both work for, or if you want to call it another name why not use VMDG480 and the will be no need for the moderators to edit posts.

The problem is the use of the word 'super' by your employer's marketing dept. I'll accept the word 'hub' as a loose description for a combined modem/router device - cable or ADSL. Let's leave aside the various reported or alleged problems with it's implementation.

Just on it's specification - what is 'super' about the VMDG480?

[Nopanic]

Originally Posted by zekeisaszekedoes

This is sorta what I meant, yeah.

Also:



Because there's a good chance that with the proper testing their superhubs would choke in the same way everyone elses does. For example, Masque never did get back to me with the results of a "transfer 5GB+ file from 300Mbps wireless N device to 1Gbps ethernet device" test, which used to break mine all the time. Light, regular usage doesn't always show up the problems it has, perhaps leading to the misinformed conclusion that their superhubs are perfect.

Trust me, even the very best kit I own, hand-picked for the job, isn't perfect and my superhub is very far from top of the heap.

Fair enough, good answer.

[qasdfdsaq]

Technically (by networking terminology standards) it's not a hub either but that's a discussion for another day.

I guess maybe the reason many people don't have problems is because
a) 300mbps (HT) mode is flaky
b) Most people probably don't only have 5-Ghz devices and leave the SH on 2.4Ghz (or don't know enough to change it)
c) HT mode should be off by default on 2.4Ghz hence HT bugs won't affect most average consumers.

[zekeisaszekedoes]

Originally Posted by qasdfdsaq

As I've mentioned in another thread, 300mbps wireless 'N' is notoriously difficult to get right, most firmware handles it OK in an ideal/greenfield situation but as soon as there's congestion/noise/interference many will fall over and cripple itself, die, or reboot. Even dd-wrt and Openwrt, considered by some as the best OSF around do this regularly. Many of the products that I know of that work "reliably" in HT + MIMO mode do so by being "bad neighbours" and ignoring the interference-mitigation sections of the wireless spec.

True enough, but running DD-WRT r16785 on the DIR-615 gives 130Mbps and average transfer speeds of 7-9MB/s, which is about what I was getting with the superhub on 300Mbps, which I consider a respectable result. I think things were faster with r14929 (more of a "bad neighbour" version), but not by much. Now I have a gigabit-capable switch if I really need to move things between the laptop and desktop at high speed I can just plug an ethernet cable in.

My contention is, if the superhub was released with use of the 100Mbps service in mind, having to set the wireless N to the more stable 145Mbps mode is going to put you well below peak bandwidth in a real-world situation (maybe 3-5MB/s if you're lucky). DD-WRT will run fine for weeks at a time even after large file transfers on a cheap router, so I don't think it's too much to ask for the "flagship" VM CPE which is supposed to replace it to do the same.

Originally Posted by Nopanic

Fair enough, good answer.

I try.

[Sephiroth]

I took 50 meg one month before that wretched SH was forced on the poor trusting peops (I accept that there is no more production of the VMNG300).

Anyway, the only involuntary reboots I've had could be traced in the event log to actions at the VM end. Very rare indeed. I keep a running event log record and stats snapshots as well as Thingbroadband graphs. So I can relate logged events to external behaviour.

My VMNG300 won't sync above 7dBmv input power (the cabinet is 70m away). I'm blessed with a number of FPAs so I can substitute them winter/summer so as to maintain something close to 0 dBmv. My Rx-MER is around the 34½ mark (was around 32½). Correctables are high, but with the improved SNR the rate of increase has halved (just the 2 or 3 dB). Uncorrectables do not rise at all. Performance is fine, although VM are beginning to load up my UBR - possibly as a swing server while they upgrade others to the new line cards; I'm not sure.

The modem connects to an Airport Extreme router which is as good as the best of the rest. Certainly 5GHz wireless performs well for those who use it simultaneously with the 2.4 GHz mode. Apart from wireless, all PCs come off Powerlines/Homeplugs which deliver the full 50 meg.

I have BT Infinity running across the same LAN - but that's a different (satisfactory too) subject.

VMNG300 is a good 'un.

[zekeisaszekedoes]

Originally Posted by Sephiroth

VMNG300 is a good 'un.

It is. Interesting post with all the technical information.

[Ignitionnet]

Resurrecting this thread I've just seen today some modem stats from a guy running on 5 downstreams, the additional one added 8MHz below the previous lowest one at 299MHz / 298.75MHz - the VMNG300 can manage 4 downstreams so an area running at 50% utilisation downstream on 5 downstreams will cause VMNG300s to run with visible contention while the Superhub punters will be fine up to 60%. Once a 6th downstream is added the Superhub is fine up to 66%, 8 downstreams 75%. 100Mbps difference.

5+ downstreams is the intended 100Mbps build, 4 downstreams being an interim solution while hardware was being upgraded.

Downstream Channels
Lock Status Modulation Channel ID Max Raw Bit Rate Frequency Power SNR Docsis/EuroDocsis locked
Locked QAM256 140 55616000 Kbits/sec 322750000 Hz -3.9 dBmV 38.8 dBHybrid
Locked QAM256 136 55616000 Kbits/sec 290750000 Hz -3.5 dBmV 38.4 dBHybrid
Locked QAM256 137 55616000 Kbits/sec 298750000 Hz -3.4 dBmV 38.6 dBHybrid
Locked QAM256 138 55616000 Kbits/sec 306750000 Hz -3.5 dBmV 38.5 dBHybrid
Locked QAM256 139 55616000 Kbits/sec 314750000 Hz -3.6 dBmV 38.3 dBHybrid
Unlocked Unknown 0 0 Ksym/sec 0 Hz 0.0 dBmV 0.0 dB Unknown
Unlocked Unknown 0 0 Ksym/sec 0 Hz 0.0 dBmV 0.0 dB Unknown
Unlocked Unknown 0 0 Ksym/sec 0 Hz 0.0 dBmV 0.0 dB Unknown

---------- Post added at 16:27 ---------- Previous post was at 16:15 ----------

I wonder if anyone remembers the VMNG becoming unresponsive under load way back when

If you were using the connection too much it would forget to respond to keepalives from the CMTS causing T3s, T4s and disconnects, it couldn't handle doing its normal work along with doing TurboDOX so they ended up having to disable TurboDOX to stabilise it.

Amusingly the Superhub I have here, in modem mode, is more reliable than my VMNG300 was during the same part of its life. Once modem mode is widely available things should be nicely sorted.

Just a little reminder. The Superhub is excrement but the VMNG300 was excrement once as well.