The Silver Surfer Thread (iBurst 2Mbps Modem)

eltherza said:
A normal 5GB wireless account (ping test taken now : http://www.pingtest.net/result/23692703.png )
Click to expand...

Just I heard about *special* accounts that are unshaped and have better QoS, than regular. Another thing I heard that there is permanent shaping (left by previous CEO's) that limit users, so they cannot degrade network any more. One more thing I heard/seen: in times of Shaun (previous iBurst representative), if users experienced problems like mine (superiorly degraded service). After couple hours, user received a phone call from iBurst technical team and asking what speeds are. User surprisingly finds that speeds were close to 80% of theoretical speed, instead of 0%-20%.

Here's mine: http://www.pingtest.net/result/23696369.png :mad:
 
Last edited:
Jagtiger said:
Just I heard about *special* accounts that are unshaped and have better QoS, than regular. Another thing I heard that there is permanent shaping (left by previous CEO's) that limit users, so they cannot degrade network any more. One more thing I heard/seen: in times of Shaun (previous iBurst representative), if users experienced problems like mine (superiorly degraded service). After couple hours, user received a phone call from iBurst technical team and asking what speeds are. User surprisingly finds that speeds were close to 80% of theoretical speed, instead of 0%-20%.

Here's mine: http://www.pingtest.net/result/23696369.png :mad:
Click to expand...

Wow, that is a socking ping! I don't have a special account [ I don't think? ]

Here is my ping now: http://www.pingtest.net/result/23718590.png My ping is generally better during the day. I've seen another iburst modem around my area, wonder if that is the cause for increased latency...
 
eltherza said:
Wow, that is a socking ping! I don't have a special account [ I don't think? ]

Here is my ping now: http://www.pingtest.net/result/23718590.png My ping is generally better during the day. I've seen another iburst modem around my area, wonder if that is the cause for increased latency...
Click to expand...

My friend (yes, he has very high speeds in my area in comparison) ordered an special account for gaming that was more expensive than regular accounts. There were prices on iBurst website, but cannot find them any more.

If you only person using iBurst in area, then it is not amazing to find such latency.
 
Jagtiger said:
My friend (yes, he has very high speeds in my area in comparison) ordered an special account for gaming that was more expensive than regular accounts. There were prices on iBurst website, but cannot find them any more.

If you only person using iBurst in area, then it is not amazing to find such latency.
Click to expand...

Well, I dont know that one for sure, but I did do tests with 2 iburst modems during that free trial period.

I posted it somewhere on the forums here, but with 1 modem I was going at 100KB/s download, I then connected with another modem (in a different part of the house) and combined speed was somewhere near 160Kb/s - 180Kb/s. The 100KB/s dropped by 10-20%.

So, in a guess, you can say that every 15KB/s below 120KB/s you go at (provided signal strength is near 100% and error rates are less than 2%) means that there is another person on your tower. But remember, people don't "just download", it dips and spikes depending on browsing/gaming/downloading habits.

From this, I get a feeling that iburst tries to keep the users per tower to less than about 8 to keep *some* quality up

ping test now: http://www.pingtest.net/result/23745480.png
 
eltherza said:
Well, I dont know that one for sure, but I did do tests with 2 iburst modems during that free trial period.

I posted it somewhere on the forums here, but with 1 modem I was going at 100KB/s download, I then connected with another modem (in a different part of the house) and combined speed was somewhere near 160Kb/s - 180Kb/s. The 100KB/s dropped by 10-20%.

So, in a guess, you can say that every 15KB/s below 120KB/s you go at (provided signal strength is near 100% and error rates are less than 2%) means that there is another person on your tower. But remember, people don't "just download", it dips and spikes depending on browsing/gaming/downloading habits.

From this, I get a feeling that iburst tries to keep the users per tower to less than about 8 to keep *some* quality up

ping test now: http://www.pingtest.net/result/23745480.png
Click to expand...

There is no certain towers here.
I ran couple of tests using iConnect, that looks inside modem and give technical info on connectivity. Results were interesting: signal levels were low and it connected very distant towers (+30km away AND no LoS :eek:) and changed to many towers, close and distant. Then it said, the best tower is that is 14km away, not one that 2km away. :eek: And indicator flashes in green. :mad:

Pingtest: http://www.pingtest.net/result/23747623.png
 
Last edited:
If you got that ping result from a tower that is 14km away, then it ain't that bad. :)

But it seems that generally your service sux. I promise I will stop complaining about mine. Still gonna switch as soon as I can.
 
SharkBait said:
If you got that ping result from a tower that is 14km away, then it ain't that bad. :)

But it seems that generally your service sux. I promise I will stop complaining about mine. Still gonna switch as soon as I can.
Click to expand...

No, it actually changes towers at very high rate (around second). And it says that 14km tower has 6dB difference (significant difference) between closest tower. It not chooses particular tower.

Pingtest right now: http://www.pingtest.net/result/23753872.png
 
Last edited:
Jagtiger said:
Just I heard about *special* accounts that are unshaped and have better QoS, than regular. Another thing I heard that there is permanent shaping (left by previous CEO's) that limit users, so they cannot degrade network any more. One more thing I heard/seen: in times of Shaun (previous iBurst representative), if users experienced problems like mine (superiorly degraded service). After couple hours, user received a phone call from iBurst technical team and asking what speeds are. User surprisingly finds that speeds were close to 80% of theoretical speed, instead of 0%-20%.

Here's mine: http://www.pingtest.net/result/23696369.png :mad:
Click to expand...

This is very serious allegation if iBurst employs such practices against their customers! r00igev@@r, can you comment on this?
 
eltherza said:
r00igev@@r hasn't logged in since sunday, perhaps he is on leave.

http://www.pingtest.net/result/23780505.png
Click to expand...

Sorry, I have been travelling and busy. The family conspired to have their birthdays all within a few days of each other!

However, the comments/allegations are false.

Dependant on the path some upstreams have better latency than others but this is out of our control (I think they call this the Internet) Additionally, the radio interface is subject to fluctuations and influenced by environmental, geographic and structural conditions.

Finally, we do shape. The Wireless 5 and above are less shaped w.r.t. P2P than other accounts (This is one of the benefits of the higher value accounts). P2P is the only shaped protocol and we do this to provide congestion avoidance on towers. P2P is aggressive in its connection count and results in poor timeslot arbitration. This is common on all wireless interfaces and we implement the policy to provide a fair service to all subscribers as the spectrum allocation is only 5Mhz. If we let P2P run rampant the service would degrade.
 
^My family also conspired to have their birthdays within a few days of each other, my poor bank account hates it.
On the topic of shaping, would it not be possible to stop shaping after hours?
 
r00igev@@r said:
Sorry, I have been travelling and busy. The family conspired to have their birthdays all within a few days of each other!

However, the comments/allegations are false.

Dependant on the path some upstreams have better latency than others but this is out of our control (I think they call this the Internet) Additionally, the radio interface is subject to fluctuations and influenced by environmental, geographic and structural conditions.

Finally, we do shape. The Wireless 5 and above are less shaped w.r.t. P2P than other accounts (This is one of the benefits of the higher value accounts). P2P is the only shaped protocol and we do this to provide congestion avoidance on towers. P2P is aggressive in its connection count and results in poor timeslot arbitration. This is common on all wireless interfaces and we implement the policy to provide a fair service to all subscribers as the spectrum allocation is only 5Mhz. If we let P2P run rampant the service would degrade.
Click to expand...

Apologises then, needed to make sure, that there are no other "hidden" factors. Please, can you comment on mysterious solution of problems I've described?
One more thing I heard/seen: in times of Shaun (previous iBurst representative), if users experienced problems like mine (superiorly degraded service). After couple hours, user received a phone call from iBurst technical team and asking what speeds are. User surprisingly finds that speeds were close to 80% of theoretical speed, instead of 0%-20%.
Click to expand...
Also, can acoustic interference from 5,5KW compressor block signals completely?


Tradition: http://www.pingtest.net/result/23804477.png
 
Last edited:
r00igev@@r said:
Finally, we do shape. The Wireless 5 and above are less shaped w.r.t. P2P than other accounts (This is one of the benefits of the higher value accounts).
Click to expand...
Would you mind explaining how the shaping is different on the Wireless 5 and above? In another thread you stated if P2P traffic was detected that the number of connections is limited to 4.
r00igev@@r said:
P2P is aggressive in its connection count and results in poor timeslot arbitration. This is common on all wireless interfaces and we implement the policy to provide a fair service to all subscribers as the spectrum allocation is only 5Mhz. If we let P2P run rampant the service would degrade.
Click to expand...
I can understand how the number of connections could affect the performance of your firewall, but how is downloading 1Mbps of HTTP traffic different to downloading 1Mbps of P2P traffic on the wireless interface, isn't it all just packets?

Vodacom don't have any P2P shaping on their 3G network, in fact my best ever experience (1.5GB in 1 hour) with BitTorrent has been on their network. Why doesn't their wireless network seem to be affected by this?
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X