I have added the --status and --pbParams info from the modem as of a few minutes ago.
Thanks for adding those, and sorry I was slow at replying.
Should the DLM have intervened?
As far as I can make out from the stats, your line is working pretty well, and there doesn't look to be much reason for FEC to be active - it certainly isn't protecting much.
I don't think DLM should have intervened - in the old-style sense of turning on high levels of FEC and interleaving. And I don't see why it will not have activated G.INP retransmission either.
As I mentioned before, your line does have FEC active on it, but it really isn't obvious why.
When FEC is on, you can see the relative performance by looking at the counts for RS, RSCorr and RSUnCorr. That tells you have many small RS blocks were received, how many had a correctable fault (RSCorr), and how many had an uncorrectable fault (RSUnCorr). If a FEC process is worth running, you'd expect to see more "RSCorr" and "RSUncorr" combined, relative to "RS". That would show errors happening. And
effective FEC process would then show RSCorr to be much larger than RSUnCorr.
In your case, you aren't getting that many errors happening (RSUnCorr + RSCorr is only 0.0004% of the RS ... a tiny proportion). That tells me that the FEC process isn't really needed.
Then RSCorr is only about the same size as RSUnCorr, which suggests the FEC process isn't actually correct much on the few occasions it is called into action. An effective FEC process would be correcting 10x, 100x or 1000x as many as are failing. One reason is that interleaving, tuned correctly, can make the FEC process considerably more efficient.
For comparison, on my line (with FEC active, alongside interleaving, alongside retransmission), I am seeing about 100x as many errors (RSUnCorr+RSCorr compared to RS), but every single one of them gets corrected. In 90 days, I have totalled 560039 RSCorrs and 0 RSUnCorrs.
The RSUnCorr's you are seeing - 2748 in the sample you sent - have a tendency to aggregate together within contiguous blocks. In your case, they turn into 509 errors measured in larger blocks (aka OHFerr). In time, these 509 errors are seen in 350 separate seconds - the "ES" measure, or "Errored Seconds".
The "ES" measured over 24 hours is one of the major parameters monitored by DLM. Your count doesn't ever seem to get high enough to interest DLM.
In summary:
- Your ES level isn't high enough to trigger real DLM intervention.
- Your FEC error level is tiny.
- The FEC level may be small because FEC is tuned down, and interleaving is turned off.
What brought me to really ask about the line was that I have a decent upload and the download being quite far off the estimate (even though it is a best effort service)
I don't see anything new to help explain why your speeds are as they are. All the comparisons I made before still stand: A few cumulative "not so good at X" means your downstream speed is low, while a notable "really good at Y" means your upstream speed is rather high.
I've had 3 different FTTC lines. If I compare your "pbParams" attenuations to those lines. I see:
| Line | Length | VDSL Band Status U0 U1 U2 U3 U4 D1 D2 D3 | Speed Attainable |
| Yours | ? | Line Attenuation(dB): 7.4 31.2 46.6 N/A N/A 14.7 38.0 60.1 | 38/19 |
| First | 650m | Line Attenuation(dB): 5.5 27.8 40.9 N/A 14.1 34.5 53.3 | 60/16 |
| Second | 375m | Line Attenuation(dB): 4.3 21.7 31.7 N/A 11.4 27.4 41.7 | 78/23 - 84/26 |
| Third | 100m | Line Attenuation(dB): 1.9 8.4 11.0 N/A N/A 4.6 10.4 16.9 | 107/33 |
All those attenuation figures suggest your line is longer than any of the ones I have had. Comparison with my slowest line's attainable speed (which was achieved late 2011) suggests you ought to be slower than that ... and indeed, you are in the downstream. But upstream is better.
Do you know how far from the cabinet you are? Following the line route, that is.
Separately, I'd estimate that the cabinet is about 2-2.5km from the exchange.
What estimates were you given? Are they the same as you ones you can get from this checker:
a) Put your phone number into this one:
http://dslchecker.bt.com/adsl/adslchecker.welcomeb) Put your postcode only into this one; on the next screen, select your address from the dropdown list:
http://www.dslchecker.bt.com/adsl/ADSLChecker.AddressOutput
