Does that mean you are a Linux rather than a Windows user?
It does indeed. Well spotted. I guess it Takes one to Know one, eh? ;-)
Your QLN graph does appear to confirm a 'noisy' line.
I believe -140 dBm is classed as a 'quiet' line, so your spikes to around -110 dBm do suggest some interference at certain tones/frequencies, possibly caused by commercial radio stations.
Others may be able to suggest which stations would cause spikes at those particular frequencies.
That seems like a good idea for me to improve the graph plotting with more informative x-axis.
Your Line Attenuation looks quite high (LATN(dB): 61.0 31.5), but probably as to be expected for your 4000m line length.
I am putting that figure of 61 down to mis-reporting by the Billion firmware.
IIRC it was reported as 51.1 - 51.2 on my Netgear DGN3500B and around 52.3 on HomeHub3 (both Infineon AR9)
Similarly around 51.1 on my Netgear DGN2200v3 (Broadcom)
So I am just treating that as a 'maximum reading' as it were.
Down Up
SNRM(dB): 3.3 17.0
Your US SNRM against a US sync speed of 448 Kbps (Channel), with an attainable (Max) rate of 984 Kbps, suggests that you could possibly achieve higher US sync speed.
I believe that some/most ISPs initially cap US sync speed at 448 Kbps by default, but some (e.g. Plusnet) are willing to uncap it on request.
The ISP on this line is BT.
Do they uncap upstream?
Or is this yet another reason to abandon them as soon as contract permits.
ES: 8501 0
That DS Errored Seconds count does seem very high for only 1 days 6 hours 20 min 15 sec DSL connection time.
Fortunately the SES is not too high to live with though.
Could you not grab the PC's local system time & use that in your scripts?
That's what my older Windows batch file scripts & current programs use.
I tag the data file with the time when it was collected.
I mentioned it to explain the implausible date.
- and in the hope that somebody here might know of a command for manually setting the clock!
Have you created any scripts that would record & plot graphs for stats obtained 24/7?
Sure. I scraped the status from http on a cron job every 60 seconds for the DGN3500B for months.
Then I rewrote that for the DGN2200.
I just appended a line of comma separated data to my log file each time.
But once I got the line stable at 3500, things didn't fluctuate so much. so eventually I stopped.
I have only had the Billion for 15 days so I recon I have done pretty well so far!
ATM I grep the data to pick out the essentials, generate the graphics and display a summary.
The current data files overwrite the previous.
There is such a large amount of data each time, i had not quite decided yet what to append to a log and what to discard!
Here is the summary I see each time I run the script.
UPTIME1: 14 days
SNTP date: Sat Jan 15 19:11:43 2000
Link time: 1 days 6 hours 20 min 15 sec
RETRAIN_REASON1: 1
RETRAIN_STRING: kAdslPhysStatusNoDefect -
SNRM(dB): 3.3 17.0
LATN(dB): 61.0 31.5
TxPwr(dBm): 19.4 12.5
RATE_SUMMARY: 5184 448 5760 984
SESSION_START: 2014-07-14 11:12:18
ADSL_TRAINING_MARGIN1: 65444
Your graphs are formatted slightly differently to those prduced by my Windows programs.
Could you possibly zip the log(s) you have used so that I can attempt to plot the data in a more familiar format for me?
As above - I haven't actually started keeping a proper log of the data I have extracted yet. :-(
But I can certainly zip the full current set.
As you are seeing DS FEC errors, it suggests that DS Interleaving is ON at some level & as the value is 1, it is OFF for US.
As you would expect, interleaving is on.
hence the
'Channel: INTR'
What data do you get from the adsl info --stats command?
The D: value would confirm the depth of interleaving & INP & delay would confirm other related details.
I didn't want to post too much all at once, for my first post.
I should have guessed that to post more data is always better! ;-)
I can attach it now.
Just for comparison purposes, I have attached a snapshot montage from another G.DMT connection, but unfortunately, I can't recall that user's line length.
Thankyou. I shall pore over that later. Always good to see some elses data.
[edit:] I had some trouble trying to post with 4 .txt file attachements.
Fortunately posting my reply then adding a zip file afterwards has sorted that out. :-)
This zip archive contains the raw data.out. All the other files are created from it.
Author
Topic: ADSL1 line stats Graphs - help interpreting QLN appreciated (Read 7401 times)