Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[kitz]

I've decided to try the latest vdsl firmware blob I could find (apparently with vectoring support): gained another 2Mb

Nice improvement there
Id already put your figures into the table and not updating because the latest increase is more to do with firmware rather than any improvements from g.inp itself.

There seems to still be the Bearer0/Bearer1 issue, showing some from 0 and some from 1. 

One thing I've noticed, bitswap is not enabled on upload since G.INP, I'm pretty sure it was ON before that. Does it matter?

From my latest stats

Code: [Select]

Capability:
        bitswap         On
        sra             On
        trellis         On



Bitswap:        483/484         0/0

Indicating bitswap is being undertaken on for up and down over bearer 0 and not on bearer 1

[kitz]

PS
rp00 can you do me a favour please and check your IPprofile against your sync speed as it would be interesting to see what the result is for someone on a lantiq chipset.

Im curious about this

Calculated as 96.79% of the sync speed or 96.69% for g.inp lines.

I know theres mention of this being used to check if g.inp is active or not - but Im not convinced when it comes to say the HH5A.
There's something weird going on with sync speeds of some of the Lantiqs which I cant fathom out yet.   The speed jumps on say the HH5As are weird - I am also very unsure as some are quoting that its a definite way to find if the HH5A is g.inp capable or not.

According to Openreach and Ian Lawrence..  they should be capable in downstream direction only.   Openreach states that if a modem isnt at least capable in the downstream direction its not supposed to sync.    This is why they put out new f/w for the eci modems.   

I'm curious about atkinsong's observation about 5% difference and perhaps being the lack of RS overhead.
I really dont know - just chucking thoughts around and it would be interesting to see your results on eci but with custom f/w

[Dray]

According to Openreach and Ian Lawrence..  they should be capable in downstream direction only.   Openreach states that if a modem isnt at least capable in the downstream direction its not supposed to sync.    This is why they put out new f/w for the eci modems.   

I've never seen any reference to new firmware for the ECI modems

[rp00]

PS
rp00 can you do me a favour please and check your IPprofile against your sync speed as it would be interesting to see what the result is for someone on a lantiq chipset.

Im curious about this

Calculated as 96.79% of the sync speed or 96.69% for g.inp lines.

Sure, I'll do it as soon as I get home.

here's what Zen customer panel is reporting:
(for some reason, my initial G.INP resync is not logged)

[atkinsong]

PS
rp00 can you do me a favour please and check your IPprofile against your sync speed as it would be interesting to see what the result is for someone on a lantiq chipset.

Im curious about this

Calculated as 96.79% of the sync speed or 96.69% for g.inp lines.

Thought I'd jump in with my ratio as well Kitz - sync is 52838, BRAS is 51.08. I reckon that's pretty damned close to 96.69%.

I'm also puzzled regarding the big improvement with Lantiq v Broadcomm with G.INP applied. It's as if the Lantiq implementation of G.INP favours chip matching!

[skyeci]

Might have to give my unlocked eci modem a bash and see what happens. Maybe my cabinet would prefer it ??

[S.Stephenson]

Might have to give my unlocked eci modem a bash and see what happens. Maybe my cabinet would prefer it ??

I gain 3 mbps on a Lantiq based modem vs HG612.

[skyeci]

Might have to give my unlocked eci modem a bash and see what happens. Maybe my cabinet would prefer it ??

I gain 3 mbps on a Lantiq based modem vs HG612.

Did your latency improve out of interest? - mine was worse before a change of pair and a reset... g.inp still switched off at the mo. Latency is my issue - happy to lose some speed if I get a better latency..

thanks

[S.Stephenson]

Might have to give my unlocked eci modem a bash and see what happens. Maybe my cabinet would prefer it ??

I gain 3 mbps on a Lantiq based modem vs HG612.

Did your latency improve out of interest? - mine was worse before a change of pair and a reset... g.inp still switched off at the mo. Latency is my issue - happy to lose some speed if I get a better latency..

thanks

Ive never had interleaving so my latency is the same, it's a shame about the lack of stats though, if I notice problems I figure I'll switch back to the HG612 to see what's happening.
L

[skyeci]

Might have to give my unlocked eci modem a bash and see what happens. Maybe my cabinet would prefer it ??

I gain 3 mbps on a Lantiq based modem vs HG612.

Did your latency improve out of interest? - mine was worse before a change of pair and a reset... g.inp still switched off at the mo. Latency is my issue - happy to lose some speed if I get a better latency..

thanks

Ive never had interleaving so my latency is the same, it's a shame about the lack of stats though, if I notice problems I figure I'll switch back to the HG612 to see what's happening.
L

well I have managed to get munin setup on my monitoring pc which Jordan got going so I will be able to get some graphs out of it. Crunch time really as to whether to plug it in or not 

[rp00]

PS
rp00 can you do me a favour please and check your IPprofile against your sync speed as it would be interesting to see what the result is for someone on a lantiq chipset.

Im curious about this

Calculated as 96.79% of the sync speed or 96.69% for g.inp lines.

Due to my stupid mistake, I've resynced the line before I could run IP profile test.

New sync speed and IP profile:
Actual Data Rate:       64.068 Mb/s / 18.175 Mb/s
IP Profile for your line is - 62.01 Mbps

which equals to 96.7877

I've double-checked if R and N values are reported correctly by the script and it appears they are.

root@eci:~# dsl_cpe_pipe.sh g997fpsg 0 1
nReturn=0 nChannel=0 nDirection=1 nNFEC=32 nRFEC=16 nLSYMB=16 nINTLVDEPTH=1 nINTLVBLOCK=32 nLPATH=0

root@eci:~# dsl_cpe_pipe.sh g997fpsg 0 0
nReturn=0 nChannel=0 nDirection=0 nNFEC=255 nRFEC=16 nLSYMB=4872 nINTLVDEPTH=1 nINTLVBLOCK=255 nLPATH=0

[S.Stephenson]

So you get a ratio of 96.79%, does G.INP show as active on your status???

If it does that shows that G.INP works on Lantiq at a 96.79% ratio.

[wj66]

Hi not to clued up on vdsl2 Broadband so please excuse my ignorance, I have been following with interest G.INP on ECI Cabinets forum.
I checked the stats this morning and again this evening and noticed that the sync rates had changed but the modem doesn’t appear to have re synced would this be G.INP on my line, if I rebooted the modem would it improve my download speed, I use Vigor 130 modem to a Netgear router on ECI Cabinet, approx 350 mts from cabinet.
Thanks John

Date                                     23-Mar                   23-Mar   
Up Time                             752:15:28      846:23:57   
Firmware                             FW 3.7.8.3      FW 3.7.8.3   
DS Actual Rate                     48996000           51823000   
DS Attainable Rate               49506624           52187600   
US Actual Rate                     17529000      17465000   
US Attainable Rate             17153036      17505600   
DS Path Mode                     Fast              Fast   
US Path Mode                     Fast              Fast   
DS Interleave Depth             1                       1   
US Interleave Depth             1                        1   
Cur SNR Margin                     5 dB              5 dB   
Far SNR Margin                     6 dB              6 dB   
NE Current Attenuation    14                      15   
Far Current Attenuation   13 db              13 db   
DS actual PSD                   6.9 dB              6.9 dB   
US actual PSD                   13. 7  dB              13. 7  dB   
Xdsl Reset Times           0                      0   
Xdsl Link  Times            1                      2   
                            Near End   Far End   Near End   Far End
Trellis                         1      1                1            1
Bitswap                         1      1                0            1
ReTxEnable                 0      0                0          1
VirtualNoise                 0      0                0            0
20BitSupport                 0      0                0            0
LatencyPath                 0      0                0            0
LOS                                 0      0                2            0
LOF                                  0      0                0            0
LPR                                 0      0                0            0
LOM                                 0      0                0            0
SosSuccess                  0      0                0            0
NCD                                 0      0                0            0
LCD                                  0      0                0            0
FECS (seconds)            12704118    1172647        0        1176429
ES (seconds)            2240    38293        0        38367
SES (seconds)            0            1644        0        1647
LOSS (seconds)            0            303                0        303
UAS (seconds)            34            110195        34     110235
HECError                    0            0                0         0
CRC                            2658    58665        0         58804
RsCorrection            0            0                0         0
INP (symbols)            0            0                0         235
InterleaveDelay     0            0                0         27
NFEC                    255            255                255    32
RFEC                    16            16                16      16
LSYMB                   4700            13130        4683  16
INTLVBLOCK           255            255                255    32
AELEM                   0            ----                0          ----
sorry for the poor formatting

[rp00]

So you get a ratio of 96.79%, does G.INP show as active on your status???

If it does that shows that G.INP works on Lantiq at a 96.79% ratio.

it does!

root@eci:~# dsl_cpe_pipe.sh lfsg 0
nReturn=0 nDirection=0 bTrellisEnable=1 bBitswapEnable=0 bReTxEnable=0 bVirtualNoiseSupport=0 b20BitSupport=0

root@eci:~# dsl_cpe_pipe.sh lfsg 1
nReturn=0 nDirection=1 bTrellisEnable=1 bBitswapEnable=1 bReTxEnable=1 bVirtualNoiseSupport=0 b20BitSupport=0

[ejs]

There are lots of dsl_cpe_pipe.sh commands to experiment with.

For G.INP / ReTx specifically, there are:
rtsg - ReTxStatisticsGet - takes one number, 0=near end, 1=far end
pmrtctg - PM_ReTxCountersTotalGet - takes one number, 0=near end, 1=far end

bitswap stats:
osg - OlrStatisticsGet - takes one number, 0=upstream, 1=downstream

There's fpsg which is like g997fpsg but gives some of the other framing numbers. For some of the commands that require two numbers, perhaps varying the first number, which specifies the "channel", might give the figures for different things? I don't think anything besides 0 works, but it might be worth trying 1 or 2 anyway.