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: VDSL2 Connection Stats  (Read 7517 times)

Jaggies

  • Member
  • **
  • Posts: 92
VDSL2 Connection Stats
« on: July 04, 2014, 08:40:15 PM »

I'm not sure if this is the correct forum, but I'm happy for it to be moved to another one if it isn't.

As a new user of FTTC, and the HG612 plus the monitoring software, I don't actually know how to interpret the stats and graphs produced. I can see that my downstream synch is a lot lower than both the attainable rate, and the estimated speed given to me on sign-up. I'm still happy with the overall performance, and I'm not really hammering the connection, however it would help if I understood the graphs before I potentially raise a fault with my ISP.

Snapshot graph from a few minutes ago is attached.
« Last Edit: July 04, 2014, 10:05:17 PM by Bald_Eagle1 »
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: Connection Stats
« Reply #1 on: July 04, 2014, 09:51:26 PM »

Hi Jaggies,


I moved your original message to this new thread as it's worth a thread of its own.

I can see from your version number & the fact that parts of your Bit loading graph are coloured blue (shared tones) that you are still using the 'older' HG612 modem firmware version.

The updated version seems to improve stability & slightly improve sync speeds for some users.

The updated version also reports some of the stats differently (perhaps more accurately) & it also allocates the tones used for the band plans slightly differently.

The tones being used for your Discovery & Medley phase band plans confirm that you are connected to a Huawei cabinet DSLAM.




I would recommend that you update the HG612's firmware from one of the GUI versions available in the Experimental area accessible via this link:-

https://mega.co.nz/#F!LdJFDIJL!e_E1twsIg2kTet8mPjrb4w



My personal preference is to view the snapshot graph montage in portrait format.


Have you been obtaining the Ongoing stats?

If so, a look at your line_stats__FULL_MONTY_P-20140704-xxxx.png montage may be useful to campare against mine.


I could then try to explain the differences that may assist in understanding what each graph depicts.


I have attached my recent snapshot & ongoing montages for reference.



« Last Edit: July 04, 2014, 09:54:53 PM by Bald_Eagle1 »
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: Connection Stats
« Reply #2 on: July 04, 2014, 10:04:51 PM »

Just as an exteme example of what's good & bad, I have attached Hlog graphs from a connection I have been recently monitoring remotely.


The graph with the pronounced 'V' shape depicts a bridged tap effect (it was found to be a star wired connection -  a typical cause of a bridged tap which has the effect of reducing sync speeds).


The wiring was sorted out today & the effect in the graphs is pretty obvious.

Sync speeds immediately went from a very poor 15 Mbps to 20 Mbps.

As SNRM immediately went from 6 dB to around 9.6 dB, it suggests the connection still has more to offer.

The new sync speed of 19998 Kbps more or less confirms that DLM had 'banded' the connection to a maximum of 20 Mbps sync speed.

It is hoped that DLM will relent within a few days & allow higher sync speed.


Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: VDSL2 Connection Stats
« Reply #3 on: July 04, 2014, 10:48:20 PM »

As I have also been privy to the data, both before and after the remedial work was performed by a certain wheelbarrow trundler, I have attached (below) the corresponding bitloading/SNR v tone graphs. The improvement is clear to see.  :) 

(I just hope that my tail remains un-pecked . . .  :(  )
« Last Edit: July 04, 2014, 10:51:22 PM by burakkucat »
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #4 on: July 05, 2014, 01:23:02 AM »

Thanks for that, I've now updated to bcm96368MVWG_fs_kernel_HG612V100R001C01B030SP06_unlockedgui-nobtagent and restarted Stats Viewer. Snapshot graphs from after the upgrade and Full Monty stats from before attached.

Hope you can make something of these - it's all Greek to me...  8)
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: VDSL2 Connection Stats
« Reply #5 on: July 05, 2014, 09:34:15 AM »

It looks like your firmware update hasn't fully completed.


I would suggest you should try again, ensuring the reset button is held down for 10 to 15 seconds when powering up the modem.
It is easy to accidentally release pressure on it & only achieve a 'partial' firmware update.

This is Wolfy's updated firmware version:-

xdslcmd --version
xdslcmd version 1.0
DSL PHY: AnnexA version - A2pv6C038m.d24j
******* Pass *********


Also your band plan tones haven't changed yet, there are still blue 'shared' tones showing in your bitloading graph.

This is how the pbParams data should look with the updated firmware version:-

Code: [Select]

Discovery Phase (Initial) Band Plan
US: (7,32) (871,1205) (1972,2782)
DS: (33,859) (1216,1961) (2793,3970)
Medley Phase (Final) Band Plan
US: (7,32) (871,1190)
DS: (33,859) (1216,1852)
  VDSL Port Details   Upstream   Downstream
Attainable Net Data Rate:      3855 kbps     21412 kbps
Actual Aggregate Tx Power:        6.9 dBm      12.4 dBm
====================================================================================
  VDSL Band Status U0 U1 U2 U3 U4 D1 D2 D3
  Line Attenuation(dB): 8.3 55.7   N/A   N/A   N/A 22.1 68.4   N/A
Signal Attenuation(dB): 8.3 55.0   N/A   N/A   N/A 31.4 68.4   N/A
SNR Margin(dB): 6.2 6.2   N/A   N/A   N/A 6.0 6.0   N/A
TX Power(dBm): 0.8 5.8   N/A   N/A   N/A 11.3 6.1   N/A


Also, the text shown in red should be present when the firmware has been fully updated:-

xdslcmd info --pbParams
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason:   1
Last initialization procedure status:   0
Max:   Upstream rate = 3855 Kbps, Downstream rate = 21412 Kbps
Bearer:   0, Upstream rate = 3866 Kbps, Downstream rate = 21698 Kbps


The differences between our connection stats are visually obvious (I'm around 1100m from the cabinet so I'll never see great speeds), but I'll try to explain some of the more relvant stats:-

Discovery Phase (Initial) Band Plan
US: (7,32) (871,1205) (1972,2782)
DS: (33,859) (1216,1961) (2793,3970)
Medley Phase (Final) Band Plan
US: (7,32) (871,1190)
DS: (33,859) (1216,1852)


My connection 'Discovered' the relevant VDSL2 band plan tones, but due to distance from the cabinet, attenuation & possibly noise/crosstalk are too high for it to actually use all of them at 'Medley' phase.

The 'older' firmware version doesn't identify the current band plan tones correctly & doesn't distinguish between DS Line & Signal attenuation.

The higher Signal attenuation in my D1 band from the updated version indicates a loss of quite a bit of signal due to noise/crosstalk:-

Code: [Select]
  VDSL Band Status U0 U1 U2 U3 U4 D1 D2 D3
  Line Attenuation(dB): 8.3 55.7   N/A   N/A   N/A 22.1 68.4   N/A
Signal Attenuation(dB): 8.3 55.0   N/A   N/A   N/A 31.4 68.4   N/A

For my connection, that equates to a loss of around 10 Mbps from when it was able to sync at 32Mbps upward.

Minimum Target SNRM for VDSL2 connections is 6dB.

As our connections are in sync at not much above 6dB, it confirms that currently we are unlikely to be able to achieve higher sync speeds.

Users with SNRM of around 9dB or more could potentially achieve higher sync speeds (unless they have intentionally restricted sync speeds or DLM or the ISP has temporarily or permanently capped sync speed).
e.g. I have recently seen a connection with 9.6dB SNRM but sync speed had been capped due to a bridged tap effect at 19998 Kbps, with an attainable rate in excess of 26000 Kbps.


The D: data confirms DS & US Interleaving depth (1 = OFF i.e. fastpath):-
D:      1      1


You have a fairly high DS depth of 863 which will currently be restricting sync speeds.
ISPs have no control over Interleaving depths for VDSL2 connections, so that value will have been determined by DLM based upon error counts, number of resyncs etc. that it has seen.

If your connection was able to regularly sync at higher speeds, it would indicate that either the connection's physical properties have deteriorated or that noise/crosstalk has increased.


With Interleaving applied, Impulse Noise Protection (INP) & delay (Ping times) will have been added at varying levels.

AS my connection is currently operating on fastpath, these values are zero:-

INP:      0.00      0.00
delay:      0      0


When my connection had DS Interleaving set at a low depth of 371, these were applied at the first stage:-

INP:      3.00      0.00
delay:      8      0

As the need for deeper interleaving depth increases, these values will also increase.
I'm nott 100% sure when DLM makes the decision to increase interleaving etc.
Examining your data in the raw xdslcmd info --stats section of a Plink log will confirm at what level yours have been set.

Too high levels of errored seconds, CRC, HEC or RSUnCorr errors or too many resyncs in fairly quick succession will drive DLM to apply interleaving.

Fast path connections do not generally use FEC (Forward Error Connection) a.k.a. RSCorr or since the firmware was updated they sometimes use very low levels.

Interleaving & FEC can be good things as they drastically reduce errors that would mean data has to be retransmitted & can actually speed up actual throughput, despite sync speed being lower.

However, as gamers tend to require very low ping times to avoid being killed in an online game etc.
High levels of delay & INP as applied with interleaving aren't very helpful.

It is often 'better' for gamers to have lower sync speeds & thus lower error counts that tend to maintain fastpath connections.


Here endeth lesson #1.

I can't explain ALL the other stats, but let me know if you have any other queries.
I'll do my best to explain wherever I can.

I think I've seen your name in the Plusnet and/or TBB forums.
WWWombat occasionally visits those forums & he could give you a very detailed technical explanation behind all the stats, but the above might help with a basic understanding for now.


See the attached ongoing montage for how the error counts have changed since my connection reverted to fastpath.

TBH, I expect it won't be too long before interleaving is applied by DLM again.


Logged

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #6 on: July 05, 2014, 01:15:24 PM »

Right - my fault the firmware upgrade didn't complete, I just logged in to the GUI via my browser and used the option within the GUI to upgrade it, I didn't do the factory reset first.

I'll have another go later, but I'm off out for lunch just now.

And yes, you will have seen me on the PN forums.  8)
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: VDSL2 Connection Stats
« Reply #7 on: July 05, 2014, 03:56:00 PM »

This has been said (by me and A.N.Other) before now but it seems to have been forgotten . . .

There are two separate ways of upgrading the firmware on a Huawei HG6xx device --
  • Via the Broadcom boot-loader.
  • Via the GUI.
Each of those methods requires the firmware to be in a different format.

All available firmware images are in the Broadcom boot-loader format.

Any attempt to perform a firmware upgrade via the GUI will fail. There will be no indication that the upgrade has failed as the incorrect format firmware image is just silently ignored.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #8 on: July 05, 2014, 06:51:45 PM »

@burakkucat - thanks for that. Now done properly and graphs showing somewhat differently. This is what it looks like on resynch. Don't have a Full Monty graph yet, but once it's been back on for a bit longer we'll see...
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: VDSL2 Connection Stats
« Reply #9 on: July 05, 2014, 07:32:05 PM »

@burakkucat - thanks for that. Now done properly and graphs showing somewhat differently. This is what it looks like on resynch. Don't have a Full Monty graph yet, but once it's been back on for a bit longer we'll see...

:thumbs:  Aim for a full 24 hours worth of data capturing before producing the new graphs.  ;)

To me, your snapshot graphs indicate a good line. But then I do not have an eagle eye . . .
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #10 on: July 07, 2014, 12:20:29 AM »

Hokeydokey, here's the graphs after 29½ hours...

Just looking for guidance on why the downstream synch rate is so far below the maximum attainable rate, and if the FEC error count is excessive at almost 180 million.

Thanks!
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: VDSL2 Connection Stats
« Reply #11 on: July 07, 2014, 01:02:21 AM »

b*cat attempts to "whistle up" an analytical Eagle (type bald).
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #12 on: July 10, 2014, 06:54:52 PM »

Obviously a busy chap...

I might just take this question over to the PN forum - perhaps one of the DCT might be persuaded to run a line test for me...  8)
Logged

burakkucat

  • Respected
  • Senior Kitizen
  • *
  • Posts: 38300
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: VDSL2 Connection Stats
« Reply #13 on: July 10, 2014, 07:07:16 PM »

Using my inbuilt climbing spikes, I proceeded up the tree that was nearest to the Eagle's Nest and vocalised . . .  ;)

There was some flapping about, followed by the ejection of a small quantity of feathers. I didn't stay around to see what happened next!  :no:
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Jaggies

  • Member
  • **
  • Posts: 92
Re: VDSL2 Connection Stats
« Reply #14 on: July 10, 2014, 07:44:23 PM »

No problem - I know that anything B_E does on here and elsewhere is voluntary, and presumably we users take up considerable amounts of valuable time which could be better spent elsewhere.

I've started a thread on the PN forums, but I'll keep an eye open here as well. Up-to-date stats can be provided on request...  :)
Logged
Pages: [1] 2
 

anything