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:

Author Topic: FTTC line statistics  (Read 2986 times)

Ixel

  • Kitizen
  • ****
  • Posts: 1282
FTTC line statistics
« on: September 02, 2012, 08:42:43 PM »

Hi all,
I've attached three graphs covering the last 24 hours, last 8 days and last 36 days. I was wondering if people could give their opinion, most importantly on the anomoly of RSCorr's massive spikes. I have had speed drops a couple of weeks ago, followed by rises in speed, coinciding with an Openreach van at the cabinet each time they happened (most likely installing Infinity for someone).

Anyway, comments welcome!

Thanks.
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC line statistics
« Reply #1 on: September 02, 2012, 10:26:57 PM »

Hi Ixel,

Firstly, was the most recent resync of 28th August initiated by you or via DLM?

It looks as though your DS sync speed is indeed being restricted by DS Interleaving being at a fairly high depth of almost 1200.
Although, TBH, 74Mb or so ain't all that bad really.

This was no doubt set by DLM after seeing quite a few general errors.

The high amount of RSCorr errors more or less means that DLM is actually doing quite a good job in Correcting the errors rather than having to re-transmit them as UnCorrected errors.

I would also imagine that INP & delay (from a snapshot "Plink" log) are set at values other than ZERO.

The "messiness" as shown between 6th & 25th July could well be due to engineers accidentally "disturbing" your connection while connecting new FTTC users etc. or carrying out "other" repair works for a whole host of unknown reasons.
However, the increase in Interleaving to a depth of 2000 or so does suggest it may have been specific to your own connection.

There is a slight glitch in the graphing script as batch files have an integer maximum limit for calculating cumulative & per minute errors. Hence the appearance of a sudden starting & stopping of RSCorr errors.
They are reset to ZERO by the modem anyway once the its own limit has been reached.
This issue will hopefully be addressed in the next (fairly imminent) release of the scripts, along with more graphable data such as US error counts, Error Seconds per minute etc.


HTH in some way, but the main thing to consider is whether or not you feel you are seeing a "good" or "poor" internet experience for your personal use requirements & whether or not you see it as "better than your previous ADSL connection.

My connection isn't fantastic in FTTC terms - at around 28Mb or so sync speed, but it is noticeably "better" than my previous up to 1Mb & rather flaky ADSL connection (on a good day).

Cheers,

Paul.
Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: FTTC line statistics
« Reply #2 on: September 02, 2012, 10:44:10 PM »

I see. Thanks for the detailed explaination and reply :).

Your explaination of RSCorr, considering I am a form of programmer/freelancer, makes complete sense. My experience on the connection is excellent, obviously now that my connection isn't where it was with a 2000~ interleaving depth a few weeks ago when my connection was disturbed. Considering my original estimate was 58.5Mbps down and 18Mbps up I am pleased with my current speeds. Although I hate interleaving, compared to the ADSL days anyway, VDSL doesn't seem to have such an impact on pings with interleaving. Although I remember when I had a depth of 1 (first few days of being connected to Infinity), and the faceplate that BT installed was faulty (which I replaced as they're cheap anyway), I remember having a ping of around 9ms when performing a few speedtest.net tests. Still, my pings to date remain around 16ms-18ms so that's fine by me. If I had the choice I would rather lose a bit of download speed (higher SNRM target) as opposed to having any increased interleaving depth, delay or impulse noise protection, but obviously with DLM there's no choice given, it decides for everyone.

On my connection I successfully host eight Team Fortress 2 MvM servers, as I'm a business user with five static IP's, so I'm making good use of the available upload speed.

I look forward to your future releases of the scripts, and if you require a tester or some help (if it's programming related that is) then feel free to contact me. Thanks for the help once again.

EDIT: Oh, forgot to answer your question. The very recent disconnect/reconnect was by me, the previous one was by DLM.
« Last Edit: September 02, 2012, 10:49:08 PM by Ixel »
Logged

Bald_Eagle1

  • Helpful
  • Kitizen
  • *
  • Posts: 2721
Re: FTTC line statistics
« Reply #3 on: September 06, 2012, 05:35:17 PM »


I look forward to your future releases of the scripts, and if you require a tester or some help (if it's programming related that is) then feel free to contact me. Thanks for the help once again.


Once the compiled but basic 'C' code mentioned in another thread is fully tested, I/we may be looking for someone to add a nice looking Windows/Linux GUI to control it all.

Would that be anything you could assist with?

Logged

Ixel

  • Kitizen
  • ****
  • Posts: 1282
Re: FTTC line statistics
« Reply #4 on: September 06, 2012, 05:58:48 PM »


I look forward to your future releases of the scripts, and if you require a tester or some help (if it's programming related that is) then feel free to contact me. Thanks for the help once again.


Once the compiled but basic 'C' code mentioned in another thread is fully tested, I/we may be looking for someone to add a nice looking Windows/Linux GUI to control it all.

Would that be anything you could assist with?

I used to do GUI's, but that was in the .NET IDE (Visual Studio for example), and Java. Anything outside Java or .NET I just do plain code so I may not be able to help you in that event given that it's C, unless the GUI can be designed in another language such as C#.NET or Java.
Logged
 

anything