Kitz Forum
Broadband Related => Router Monitoring Software => Topic started by: ktz392837 on March 24, 2016, 09:26:29 PM
-
General question, if we would like to see some kind of mdws support for hacked eci modems what would need to happen? Is there anything us users could do to help? As ginp is rolling out to eci cabs and giving better sync speeds using an eci modem some mdws support is wished :)
You'd need to ask in the respective threads for HG612 Modem Stats (http://forum.kitz.co.uk/index.php/topic,15483.0.html) and DSL Stats (http://forum.kitz.co.uk/index.php/topic,17106.0.html) really, as it's those two programs which harvest the stats and upload them to MDWS.
In theory if you can telnet into the ECI modem then it should be possible, but I suppose it depends on how much work it'll take and the time available to the authors.
-
Realistically, very little chance for DSLstats, I'm afraid. It would be too big a job for me to take on now. Sorry.
-
I'm going to take a look into this, initially testing things via user skyeci to get the ball rolling
-
I'm going to take a look into this, initially testing things via user skyeci to get the ball rolling
Yes it has only been 3 days but anything to report? I am going to break (sorry solder) my ECI today so cross fingers.
Unfortunately I won't be able to test until ginp is enabled on my line as I want stats when it happens as with my luck I will probably lose speed so keeping hg612 connected.
Thanks for looking at this I hope it is possible.
-
See the attached log, (obtained automatically via a Windows batch file script for now), showing some of the stats obtainable.
It all needs reformatting for graphing & it appears that some data is averaged into groups of 8 tones, which won't be as detailed as the stats we are familiar with.
-
You could try g997sang instead of g997dsnrg. g997lspbg (G997_LineStatusPerBandGet) may also be useful for the LATN, SATN, and SNRM per band.
-
TBH, the whole stats harvesting seems rather flaky.
It seems to work for a while, then for no apparent reason, all stats harvesting ceases, apart from data from this command:-
/etc/init.d/dsl_control lucistat
Even this command stops working:-
/etc/init.d/dsl_control status
Running the commands either via the script or directly via a ssh session doesn't return any data.
The only way to get the other stats back seems to be to restart the modem.
-
The /etc/init.d/dsl_control lucistat uses a different pipe than the other script functions, so something must have gone wrong with pipe 0. The dsl_control script relies on the same dsl pipe commands for the pretty printed output.
I will add a comment on the "OpenWRT for the ECI modem" thread because it might be more reliable if there were an extra utility in the firmware (a dsl_cpe_pipe.sh command, which is no longer actually a shell script).
-
I've changed the topic title to be more appropriate.
-
Attached is the first graph generated via HG612 Modem Stats graphing program using data obtained via an ECI modem on my connection to a Huawei cabinet.
It contains DS data only for now & I'm still a long way from a finished product.
It is, however, some progress in the right direction.
Watch this space............................
-
:thumbs:
Well done, its a good start :clap:
-
Is it too soon to say when will it support mdws and rpi? Only joking looks very promising thanks for looking at this.
-
Is it too soon to say when will it support mdws and rpi? Only joking looks very promising thanks for looking at this.
I'm not sure yet, but it's getting closer, bit by bit.
I still don't fully understand what data can be gleaned from an ECI modem & some of the data that I can obtain isn't very detailed. e.g. QLN, Hlog & SNR have average values for 8 tones grouped together, unlike a HG612 etc. that provides data per tone.
As my existing programs use the familiar Discovery Phase band plans for various matters, including sorting DS & US data into differing colours, the band plan data I can get from the ECI for my connection isn't particularly helpful.
Band Plan Data from HG612 for my connection to a Huawei DSLAM:-
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) (871,1200)
DS: (33,859) (1216,1766)
Band Plan data from an ECI modem (same connection):-
nReturn=0 nDirection=0 nNumData=2
nFormat=(nBandIndex, (nLimit_firstToneIndex, nLimit_lastToneIndex), (nBorder_firstToneIndex, nBorder_lastToneIndex)) nData="
(00,(0007,0032),(0007,0031))
(01,(0871,1196),(0880,1195))
"
nReturn=0 nDirection=1 nNumData=2
nFormat=(nBandIndex, (nLimit_firstToneIndex, nLimit_lastToneIndex), (nBorder_firstToneIndex, nBorder_lastToneIndex)) nData="
(00,(0033,0859),(0048,0859))
(01,(1216,1924),(1217,1699))
"
Could someone who can achieve much higher speeds than I can please post their ECI raw band plan data via these commands:-
root@OpenWrt:~# echo "bbsg 0" > /tmp/pipe/dsl_cpe0_cmd ; cat /tmp/pipe/dsl_cpe0_ack
&
root@OpenWrt:~# echo "bbsg 1" > /tmp/pipe/dsl_cpe0_cmd ; cat /tmp/pipe/dsl_cpe0_ack
I have attached an example of graphing progress to date, but there's still a lot to do before I'll be happy to release it for Beta testing.
EDIT:
Just for cross-reference purposes, I have attached a recent example of data for my connection via a HG612 modem.
It is the example with 6 information/graph panels.
-
Did you try the g997sang command for getting the SNR per tone? I expect the reason that the some of the data has been stored as averages for 8 tones is because that's how it's required to be sent to the other end when requested (G.993.2 11.4.1 - there's a maximum of 512 groups).
-
e.g. QLN, Hlog & SNR have average values for 8 tones grouped together, unlike a HG612 etc. that provides data per tone.
Can you apply some artistic license and use the average value for all 8 tones for graphing?
I am going to have to try and get my eci soldered up not my best skill I hope eci modems are still cheap on ebay for when I solder all the 4 pins together!
-
e.g. QLN, Hlog & SNR have average values for 8 tones grouped together, unlike a HG612 etc. that provides data per tone.
Can you apply some artistic license and use the average value for all 8 tones for graphing?
It looks like it already is, although the SNR data has been stretched by a factor of 2 along the x-axis.
-
Did you try the g997sang command for getting the SNR per tone? I expect the reason that the some of the data has been stored as averages for 8 tones is because that's how it's required to be sent to the other end when requested (G.993.2 11.4.1 - there's a maximum of 512 groups).
Every time I obtain the data using g997sang there are quite a few missing groups.
Assuming that means that my connection can't actually use anything those groups, I have added the groups with a value of zero for each of the 8 tones within the group.
The resulting graph isn't at all pretty (see attached).
Now, as when using "g997dsnrg 1 1", data is reported at the higher frequency groups that wasn't previously reported via my HG612 modem.
Considering my connection can't use these frequencies due to attenuation over 1100m or so of copper from the cabinet, I am starting to think there is some sort of error regarding SNR data (regardless of which command is used to obtain it).
Data for the other graphs looks reasonably similar to graphs from my HG612 (apart from Hlog which now shows a lot more at the higher frequencies than with the HG612).
-
I thought the g997sang data wasn't grouped, but it does skip the not measured values of 255. There also g997sansg which prints the data in a more compact format, although I'm not sure how useful that would be with the different bands of VDSL2.
The formula the convert the printed values to SNR is given in the source code for the dsl_cpe_control program, it's -32 + value/2.
-
Using g997sansg for SNR & converting the hex data to decimal values, it still appears wrong when averaging 8 tones per group to make up all 4096 tones.
(See the attached individual SNR graph).
However, if I just use 4 tones per group for SNR i.e. 2048 tones in total, the output looks fairly similar to that produced via data from a HG612.
(See the attached montages).
Bitloading, QLN & Hlog graphs still use 8 tones per group in the ECI montage.
I'd still like to see the raw data from a connection that can use almost all 4096 tones (or at least some of all the available bands) for comparison purposes.
-
Also, when only using 4 tones per group for SNR & 8 tones per group for Bitloading, the Bits with SNR plots actually line up with each other.
-
I've just recieved my OpenWRT ECi modem, here's the stats.
If you need anything from me if you explain how to do it i'll give it a whirl.
Seems good to me it knocked my INP from 49 down to 45 :)
-
Where did you get it from S.Stephenson? Which model is it exactly? There seem so many!
TIA ;)
-
I sent off a ECI modem I had and got it sorted for £15. http://www.ebay.co.uk/itm/231887970610?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT (http://www.ebay.co.uk/itm/231887970610?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT)
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fs15.postimg.org%2F54iwy9uu3%2Fimage.jpg%5Dhttp%3A%2F%2Fs15.postimg.org%2F54iwy9uu3%2Fimage.jpg&hash=1e3c2567aa45963d3d0c8d498791269081dcf68e)
-
Thanks for all the info. Much appreciated.
-
Ready to test on a RPi and ECI modem when ready :)
Could we write a script that transforms ECI telnet output to HG612 format and this is used for MDWS?
-
Could we write a script that transforms ECI telnet output to HG612 format and this is used for MDWS?
Can't see why not, it's just a matter of matching data items to each other
-
Would it be possible to include more generic Lantiq support, such as for the TP-Link TD-W9980 (unlocked/hacked to make shell access available)? There's no /etc/init.d/dsl_control on the TP-Link, so you'd need to do a lot of different commands to get each item of data.
This way does work, and the output should be the same format:
echo "vig" > /tmp/pipe/dsl_cpe0_cmd ; cat /tmp/pipe/dsl_cpe0_ack
However, it might be more reliable to do it this way instead:
/firmware/dsl_cpe_pipe.sh vig
-
Any further updates to how this is going? I've switched back over to my ECI modem as it syncs at 66mbit/s on my G.INP line vs 62-63 on Broadcom, I'm hoping for 79999/20000 on that line whenever 3db goes live.
The ECI vs HG612 on my non-G.INP line is negligible.
-
To be honest I think there are plenty of third party VDSL modems out there which work straight out of the box with DSLstats & HG612_Modem_stats without having to re-write or flash the ECI modems to gain access to stats.
-
I suppose so, but the ones that can easily be monitored work best on Huawei, in my experience the ECI modem/Lantiq chipset has a very positive effect on sync rates when G.INP is active I know that at the moment G.INP isn't active on many lines, but when G.INP is active on ECI again I can see more people switching to ECI modems.
-
I suppose so, but the ones that can easily be monitored work best on Huawei, in my experience the ECI modem/Lantiq chipset has a very positive effect on sync rates
I see ECI cabinet users doing better with a broadcom modem than a lantiq modem and them I here and ECI cabinet users sees better line stats with a lantiq modem even on ADSL way back in 2007 before I understood this stuff my purchased modem/router from PC world was better than the ISP modem because it was a broadcom chipset
No Huawie cabinets or ECI cabinets in those days just Broadcom it's just a better chipset for broadband.
-
Whilst I agree I saw a higher sync when my line did have g.inp on my unlocked eci modem I switched over to an 8324 on bridge mode not long after losing g.inp, at present the 8324 has synced just below the eci with g.inp so if g.inp ever does come back I hope 8324 will sync above the eci.
"Be" has one of my eci modems which he spent some considerable time trying to make things work but I know he encountered a number of problems.
At least I have now learned how to flash the eci modem so have a few sat on the shelf now unlocked should things ever change with stats or sync improvements in the future ::)
-
I do understand this trying to match users chipset to the cabinet vendor even had to decommission my 4 year old HG612 just something about it seemed odd with the SNRm any device close by when switched on would cause the SNRm to drop suddenly by 2 or 3.5dB. :crazy:
-
Whilst I agree I saw a higher sync when my line did have g.inp on my unlocked eci modem I switched over to an 8324 on bridge mode not long after losing g.inp, at present the 8324 has synced just below the eci with g.inp so if g.inp ever does come back I hope 8324 will sync above the eci.
"Be" has one of my eci modems which he spent some considerable time trying to make things work but I know he encountered a number of problems.
At least I have now learned how to flash the eci modem so have a few sat on the shelf now unlocked should things ever change with stats or sync improvements in the future ::)
I have a 8324 and it synced at 63 vs 66 on the ECI.
Hopefully it beats the HG612 on my non G.INP enabled line, I'm just waiting for my crosstalker to go offline again so I can keep my current sync.