Superhub : Superhub 2

[GrimUpNorth]

Originally Posted by craigj2k12

So are you saying the SH2 responds to pings from TBB at lower priority, or that TBB send SH2 pings at low priority?

Thought it had, but could well be wrong - sure someone in the know will be along soon. All I can say is that I've just spent a day in router mode and my TBB was very yellow, but speedtest, pingtest and my samknows box gave similar results to modem mode.

Cheers

Grim

[SnoopZ]

Originally Posted by craigj2k12

The hard evidence shows otherwise

Seems pretty good to me though in SH2 router mode, but i am no expert, but i think Seph knows alot more about this.

[MacmillaN]

Originally Posted by SnoopZ

Yes alot of people use inSSider to scan.

http://www.metageek.net/products/inssider/

Enginner did not check any Wireless channels? Could that possible solve the issues?

[GrimUpNorth]

Originally Posted by MacmillaN

Enginner did not check any Wireless channels? Could that possible solve the issues?

It'll let you see which channels are less congested. I check regularly as I don't set the channel to automatic.

Cheers

rim

[horseman]

Originally Posted by craigj2k12

So are you saying the SH2 responds to pings from TBB at lower priority, or that TBB send SH2 pings at low priority?

He may not want to say that, but I will! However if/when Seph re-engages he'll no doubt quite correctly point out that he conversely does not currently experience that level of jitter in RouterMode.
I believe that may be explained and dependent on combination of T patch level and concurrent utilisation/load on DOCSIS chipset and thus the overhead of subsequent line conditioning on individual connections.

De-prioritising ICMP/traffic targeted specifically at a router rather than "beyond it" is after all not uncommon.

Like no doubt yourself and many others I don't rely on a single diagnostic either - I run other BQM/ping plotters in parallel in addition to SK monitors (I use the plural deliberately as perversely for a short period over last week or so I am literally running two seperate SK units in parallel! )

That's in addition to also monitoring typically a minimum of 2 or 3 other CMTS subscribers TBB/BQM monitors remotely in my Brighton cluster.

Just for your further amusement, early in May I was stress testing my first SH2(yes I'm now on my second) by saturating my upstream (that I knew had an open utilisation fault on both d/s & u/s) with following colourful result:

[Sephiroth]

By command from the Horse:

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

[progers]

Originally Posted by Sephiroth

By command from the Horse:

Mine's improving believe it or believe it not!

[SnoopZ]

Originally Posted by Sephiroth

By command from the Horse:

Do you think you have changed some settings in the SH2 that may account from your extremely rare normal looking graph?

I would be grateful if you're able to list what is enabled and disabled so i could possibly try and replicate your result.

[Sephiroth]

FIREWALL
Firewall Disabled
Firewall Protection Low
Port Scan Detection Disabled
IP Flood Detection Disabled
IPSec Pass-Through Enabled
PPTP Pass-Through Enabled
Multicast Pass-Through Enabled

No other security settings changed from default.

Wireless Band Settings
2.4GHz Enabled
5GHz Enabled

everything else default. WPS Disabled.


UPnP Disabled


If I turn fireqwall on, it makes no difference.

Good luck.

[SnoopZ]

Thanks for your help, i will change that and see what happens but i don't expect any change.

[SnoopZ]

Originally Posted by SnoopZ

Thanks for your help, i will change that and see what happens but i don't expect any change.

Well it didn't make any difference.

[Sephiroth]

Didn't expect it would. As I've said before, VM have offered me no explanation on the trial forum as to whay I have this TBB report in router mode. If there was a neighbour with a SH2 .....

[SnoopZ]

I wish they would respond as something is clearly happening, although as it appears it isn't affecting my connection i guess it isn't a priority.

My computer was switched off from around 21:45 till 1pm and it still spikes on the graph.

As soon as i turned the computer on going by the graph the maximum latency shoots right up and even reports some packet loss, however from doing a ping test i am clearly not getting any issues that i can tell.

[Kushan]

I've seen this behaviour before, it happens on some Draytek routers as well due to the ICMP requests being put on a lower priority. That's why the connection is unaffected.

Here's an example of what I mean (from my work connection, which is not VM):



Looks bad, but the connection is actually fine.

[SnoopZ]

I guess changing the ICMP requests priority would need a firmware update and having them higher or lower would that cause an issue somewhere?