Print

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

[WWWombat]

Yes, I know I intervened with DLM but this recent intervention wasn't by me, it was by DLM.

You misunderstand.

That recent DLM activity wasn't an intervention. It was one step of de-intervention.

In your haste to trigger something visible to the HH5, you probably caused a whole chain of DLM interventions that were invisible - different variants of G.INP settings - until eventually DLM chose to band your line. Banding is one of the last intervention steps that DLM chooses, not one of the first. It is one of the harshest penalties that gets applied only to serious problem lines. Once applied, DLM doesn't remove it lightly.

Now you have to wait for DLM to perform a chain of de-interventions. Some will change G.INP settings, and some will change the banding.

I'm seriously contemplating ringing BT now.

Whether this were DLM intervention or de-intervention, caused by you or the environment, the answer would be the same: tough. BT's stance is that the whole point of DLM is to intervene, to bring stability. Being allowed to reset DLM undermines this process.

Treat it as an opportunity to learn the art of patience. What would you do with that 5Mbps anyway?

I'll do a full reply for the stats later on (but, yes, the settings have reduced). For now, though, some things...

I started a reply on a different computer, but I doubt I'll get chance to complete that in a hurry. Monday is more likely, sorry.

[Chrysalis]

u only causing yourself grief checking daily your banding something that is known to stick for months

[William Grimsley]

Yes, I know I intervened with DLM but this recent intervention wasn't by me, it was by DLM.

You misunderstand.

That recent DLM activity wasn't an intervention. It was one step of de-intervention.

In your haste to trigger something visible to the HH5, you probably caused a whole chain of DLM interventions that were invisible - different variants of G.INP settings - until eventually DLM chose to band your line. Banding is one of the last intervention steps that DLM chooses, not one of the first. It is one of the harshest penalties that gets applied only to serious problem lines. Once applied, DLM doesn't remove it lightly.

Now you have to wait for DLM to perform a chain of de-interventions. Some will change G.INP settings, and some will change the banding.

I'm seriously contemplating ringing BT now.

Whether this were DLM intervention or de-intervention, caused by you or the environment, the answer would be the same: tough. BT's stance is that the whole point of DLM is to intervene, to bring stability. Being allowed to reset DLM undermines this process.

Treat it as an opportunity to learn the art of patience. What would you do with that 5Mbps anyway?

I'll do a full reply for the stats later on (but, yes, the settings have reduced). For now, though, some things...

I started a reply on a different computer, but I doubt I'll get chance to complete that in a hurry. Monday is more likely, sorry.

Ok, I do apologise. Thanks for your explanations, they really do help me learn.

[William Grimsley]

u only causing yourself grief checking daily your banding something that is known to stick for months

Months? Oh dear, yes I couldn't get a DLM reset. It's something I need to learn from.

[Dray]

I see you had a run in on the BT Community forum

[William Grimsley]

I see you had a run in on the BT Community forum

I certainly did. I don't think they like people trying to wing it. I really was only trying to get a DLM reset, but they were having none of it!

[Dray]

You need a fault and an engineer visit for that

[licquorice]

You need a fault and an engineer visit for that

Exactly

[William Grimsley]

You need a fault and an engineer visit for that

I do have a fault which happened 2 nights ago. But, no problem.

[licquorice]

You can't classify a one off burst of errors as a fault that could be investigated with an identifiable cause.

[William Grimsley]

You can't classify a one off burst of errors as a fault that could be investigated with an identifiable cause.

I suppose so.

[William Grimsley]

Just had another burst of OH Frame Errors, ES, SES and UAS and the DSL as well as the Internet light went out briefly.

Mode   VDSL2
Traffic Type   PTM
Status   Up
Link Power State   L0
Downstream   Upstream
Line Coding (Trellis)   On   On
SNR Margin (dB)   10.8   6.6
Attenuation (dB)   25.6   0.0
Output Power (dBm)   11.9   5.3
Attainable Rate (Kbps)   45065   8950
Rate (Kbps)   34999   8495
B (# of bytes in Mux Data Frame)   166   97
M (# of Mux Data Frames in an RS codeword)   1   1
T (# of Mux Data Frames in an OH sub-frame)   0   0
R (# of redundancy bytes in the RS codeword)   8   8
S (# of data symbols over which the RS code word spans)   0.1517   0.3658
L (# of bits transmitted in each data symbol)   9229   2318
D (interleaver depth)   16   8
I (interleaver block size in bytes)   175   106
N (RS codeword size)   175   106
Delay (msec)   0   0
INP (DMT symbol)   48.00   43.00
OH Frames   0   0
OH Frame Errors   1474   15
RS Words   3572160   1498389
RS Correctable Errors   42   16
RS Uncorrectable Errors   0   0
HEC Errors   0   0
OCD Errors   0   0
LCD Errors   0   0
Total Cells   9220487   0
Data Cells   269883   0
Bit Errors   0   0
Total ES   28   1
Total SES   27   0
Total UAS   137   110

[kitz]

I had 562 Downstream OH Frame Errors which caused the Downstream SNR Margin (dB) to increase further to 10 dB. How is that not a problem?

DLM is unlikely to see 562 OH Frame Errors as reason to intervene.
From your stats
Before

INP (DMT symbol)   54.00   55.00

After

INP (DMT symbol)   48.00   43.00

I'm not going to do a detailed analysis, but from the above it shows the DLM has taken a step back on error protection and reduced it.   
Because less error protection is in place, this should give you more sync speed to play with.

However your line is banded, so it cant sync at the higher speed.  The result is it shown as surplus SNRm...  in exactly the same way that a line that is restricted by a 40Mbps or 80Mbps cap will have more SNRm than 6dB.

Look at your attainable rate,  that has gone up too.   44479   8897 ->  45169   8886

DLM has reduced INP (Impulse Noise Protection).

If your error rate remains stable at the new INP parameter, then next step may be removing of the banding cap.
I think youve seen the increase in SNRm thinking DLM has increased the target SNRm,  this doesnt appear to be the case.   At the moment its the 35Mb cap which is restricting your speed.  By reducing the level of INP its given you more SNR to play with for when the cap is removed.   

[kitz]

Our above posts crossed,  I see youve just had another resync.   Let me look at the new figures now.

[kitz]

Cant see any difference in any of the DLM parameters for your last sync.

Also please don't get fixated on OH Frame errors.   ES are the really important ones when it comes to DLM.   
An increase from 12 to 28 over the course of the day is nothing.

I can only repeat what I said a few days ago that your Err Secs are exceedingly low.. or even better what wombat said

Don't get paranoid about seeing ES's or CRC's. Ignore singles, or tens of ES's per day. Monitor low hundreds. Start to worry about high hundreds. Work to improve the thousands.

When DLM eventually gets around to removing the cap and reducing INP & delay further, then your Err Secs will start to creep up.   You have been warned that it will increase.   It is not a line fault, you will just then be seeing the true condition of your line.