Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: [1] 2

Author Topic: The affects of G.INP implementation on BTW's IP profile system  (Read 6480 times)

tommy45

  • Reg Member
  • ***
  • Posts: 627

Firstly my connection has recently had G.INP enabled  by DLM, Although it syncs at the max rate of 79999 which is 13 kbps higher than when on fast path (79987)  The BTW ip profile for some reason has for some of reason varied  from a max of 77.42 if using the HG612 modem and 77.44 when using the ECI ,

Now that G.INP is enabled  my IP profile is only 77.35 why is this Should i not have a IP profile of 77.44 or 77.42 when i am syncing at the full rate,?

I have read that it is controlled by sync speed, is anyone seeing the same?
« Last Edit: March 19, 2015, 10:13:07 AM by tommy45 »
Logged

ardsar

  • Reg Member
  • ***
  • Posts: 145
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #1 on: March 19, 2015, 10:23:21 AM »

With a sync speed that high is 0.1M difference really an issue !


Sent from my iPhone using Tapatalk
Logged

tommy45

  • Reg Member
  • ***
  • Posts: 627
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #2 on: March 19, 2015, 01:10:57 PM »

By it's self, no not noticeable, But it is odd to say the least  maybe even incorrect
But when you are with an isp that has it's own IP profile  that does restrict throughput  because its set too low (77.2 ) throughput is some 1 to 1.5mbps less and this causes extra latency whilst transfareing data  because it can't max out , the same happens when congestion kicks in too
Logged

Ronski

  • Helpful
  • Kitizen
  • *
  • Posts: 4308
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #3 on: March 19, 2015, 03:42:34 PM »

Come on tommy, you're nit picking,  our connection here at work is half yours, and mine at home is not much better than our works connection.  Neither cause me any noticeable problems, would love to have a full speed connection.
Logged
Formerly restrained by ECI and ali,  now surfing along at 550/52  ;D

tommy45

  • Reg Member
  • ***
  • Posts: 627
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #4 on: March 19, 2015, 06:04:48 PM »

Call it nit picking if you wish but it is they who have caused me to become critical of the service I'm being provided for my money each month i'm sick of being slowed down in one way or another  IMO plusnot aren't doing them selves any favours with this, which  is in a way throttling  just as the  all to frequent peak time issues are ,
Logged

Chrysalis

  • Content Team
  • Addicted Kitizen
  • *
  • Posts: 7408
  • VM Gig1 - AAISP CF
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #5 on: March 19, 2015, 09:12:01 PM »

tommy how much did the attainable move?

wish I was on a hauwei cab :(
Logged

WWWombat

  • Kitizen
  • ****
  • Posts: 1674
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #6 on: March 20, 2015, 01:14:32 AM »

Remember that G.INP relies on bearer 1 using some bandwidth, for the re-transmitted packets. For some modems, I've seen it report 0.128Mbps; other modems report zero.

It seems likely, therefore, that bearer 0 - the main channel - will have lost 0.128Mbps from its throughput. Equally likely that BT's IP Profile value is calculated from the bearer 0 throughput, rather than the total sync speed. Plusnet, of course, has no influence on this.

This conclusion depends on the modem reporting a sync speed that adds together the bandwidth of bearers 0 and 1.
Logged

tommy45

  • Reg Member
  • ***
  • Posts: 627
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #7 on: March 20, 2015, 02:39:20 AM »

Well my stat's suggest        Bearer:   1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps So if it is reporting the full sync rates then why is the profile  as it is ?,

@chrys : attainable , move from when , from when interleaved 2nd step, or  prior to that on fast path? if fast path it's around 2mbit, or back to around what it was last sept or when i first was able to monitor the line , as i had a ECI modem supplied back in April 13 by BTOR  when installed so don't know what is was back then,
« Last Edit: March 20, 2015, 02:44:14 AM by tommy45 »
Logged

Ronski

  • Helpful
  • Kitizen
  • *
  • Posts: 4308
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #8 on: March 20, 2015, 06:46:12 AM »

Call it nit picking if you wish but it is they who have caused me to become critical of the service I'm being provided for my money each month i'm sick of being slowed down in one way or another  IMO plusnot aren't doing them selves any favours with this, which  is in a way throttling  just as the  all to frequent peak time issues are ,

Don't forget those of us not lucky  enough to be close enough to cab to get full speed pay exactly the same as someone who is and on the same package, so who gets the better deal? I get half what you get, and Be1 gets half what I get,  but we have the same choice of PN packages! Oh, and now there appears to be the eci/Huewie divide, so us ECI users get an inferior quality service, yet still pay the same, I would be counting myself lucky on your line.

I'm also with PN and have been since 2003, we get these congestion issues every so often, then PN buys some more capacity and the problem goes away for a while.
« Last Edit: March 20, 2015, 06:49:20 AM by Ronski »
Logged
Formerly restrained by ECI and ali,  now surfing along at 550/52  ;D

boost

  • Reg Member
  • ***
  • Posts: 768
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #9 on: March 20, 2015, 09:38:05 AM »

Remember that G.INP relies on bearer 1 using some bandwidth, for the re-transmitted packets. For some modems, I've seen it report 0.128Mbps; other modems report zero.

It seems likely, therefore, that bearer 0 - the main channel - will have lost 0.128Mbps from its throughput. Equally likely that BT's IP Profile value is calculated from the bearer 0 throughput, rather than the total sync speed. Plusnet, of course, has no influence on this.

This conclusion depends on the modem reporting a sync speed that adds together the bandwidth of bearers 0 and 1.

This! The IP profile is meant to be represent net throughput, I believe. PTM overhead is just shy of 4% minus whatever is allocated to the second latency path = a marginally different IP profile for some (all?) G.inp enabled circuits.

I think it's important to highlight stuff like this. Every little helps! :D
Logged

WWWombat

  • Kitizen
  • ****
  • Posts: 1674
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #10 on: March 20, 2015, 10:01:17 AM »

Well my stat's suggest        Bearer:   1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps So if it is reporting the full sync rates then why is the profile  as it is ?,

A better question to ask is whether the modem is telling the truth: is bearer 1 really set for zero Kbps? Or is the modem making a mistake?

If bearer 1 really was set for zero Kbps, then your modem would have no way to receive the retransmitted blocks. That patently isn't true, so the modem is getting it wrong.
Logged

Charles

  • Member
  • **
  • Posts: 19
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #11 on: March 20, 2015, 10:22:21 AM »

is anyone seeing the same?

My set up: HG612 latest firmware, Huawei cab, G.INP enabled, 79999 sync.

The BT IP Profile displayed on the speedtester went down from 77.42 Mbps to 77.35 Mbps after G.INP was enabled. No big deal.

What makes more of a difference is the Plusnet profile cap which they limit to "78 Mb" which we know restricts throughput at the top end by a couple of Mbps - http://forums.broadband-finder.info/plusnet/4315755-are-plusnet-limit-cap-on-fttc.html?fpart=all&vc=1
« Last Edit: March 20, 2015, 10:43:11 AM by Charles »
Logged

boost

  • Reg Member
  • ***
  • Posts: 768
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #12 on: March 20, 2015, 11:32:30 AM »

Well my stat's suggest        Bearer:   1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps So if it is reporting the full sync rates then why is the profile  as it is ?,

A better question to ask is whether the modem is telling the truth: is bearer 1 really set for zero Kbps? Or is the modem making a mistake?

If bearer 1 really was set for zero Kbps, then your modem would have no way to receive the retransmitted blocks. That patently isn't true, so the modem is getting it wrong.

Interesting!

Is anyone showing any unusual bitloading which might indicate where the out of band bearer is hiding? Any bins loaded that don't align with the band plan, etc?
Logged

tommy45

  • Reg Member
  • ***
  • Posts: 627
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #13 on: March 20, 2015, 02:08:09 PM »

I would be counting myself lucky on your line.
I'm also with PN and have been since 2003, we get these congestion issues every so often, then PN buys some more capacity and the problem goes away for a while.

I do count myself lucky that this line has been remained able to obtain sync at the full rate,  but it isn't without local  issues, And TBH  if i had known how FTTC would  perform  on this line prior to ordering , i would of chosen a different ISP part of my strategy when choosing isp was  what if it performs very badly ,clobbered by dlm  ect ect,  but the relatively low costs of any remaining months of the contract  was affordable, in a worst case scenario  where i should  have migrate within the  min term,

We all have different expectations of how a connection should be, in part because we all use them differently  for me a heavily interleaved connection   is a big no no,  interleave  is the enemy  low latency and good peering/routing  to me is possibly more important than the top speed ,

The ISP's congestion  problems shouldn't  be happening and when they do they take far too long to react and actually add the capacity required , this secondary IP profile , being set lower than the  bt one is saving plusnet  a lot of bandwidth, Maybe that what it's really about? I dare say it could also be akin to throttling everyone's connection,?

if you don't have several devices or people sharing your  connection then IMO you don't require their prioritisation /shaping , it is of no benefit to me,   

Logged

boost

  • Reg Member
  • ***
  • Posts: 768
Re: The affects of G.INP implementation on BTW's IP profile system
« Reply #14 on: March 20, 2015, 03:25:25 PM »

Well done for spotting that profile difference, Tommy :)
Logged
Pages: [1] 2