OK its kind of obvious now why you asked the question which didn't make sense.
Firstly DSLstats will record all errors, the reason why none were recorded after the SNRM shot up to 16dB is simply because there weren't any.
SNR Margin is a buffer zone above the true SNR. The higher the SNR Margin, then the less likely there is to be any errors. SNRM and Target SNRM is a way the DLM can help to stabilise a line. The downside for you is by increasing the Target SNRM, this takes away a large chunk of sync speed.
Have a read of:
~ What is SNR? ~ What is SNR Margin? ~ What is Target SNR Margin (TNMR)?and you will perhaps understand why the DLM is using the Target SNRM on your line.
The big question now is what is triggering the DLM to increase the Target SNRM?
There are 2 main parameters which the DLM monitors to decide if it needs to take any action
1) The amount of Errored Seconds
2) The number of line retrains (resyncs)
Looking at your line stats ~125 Err Seconds should not trigger the DLM and if you can confirm that the line didn't have a lot of disconnections the previous day then it becomes a bit puzzling why it should behave the way it is doing.
One possibility that does spring to mind - especially when your target SNRM appears to shoot straight up to ~15dB - is that TT have you on a very harsh DLM profile. There are many different profiles that TT use, but from memory 15dB is one of their maximum stability profiles.
I'm wondering if your DLM profile is too agressive. - No matter how many times TT reset the DLM, if the actual profile is too high, then DLM
could literally be kicking in just after a low rate of errors.
Try asking TT what your
DLM profile is and can it be reduced.
The profile will look something like SF_15_24M_1M. With me not being on TT I'm not sure offhand what options are available to you, but try asking if your DLM profile can be changed to a lesser one now that the line fault has been fixed by the engineer to one of the 6dB profiles. Perhaps
b*cat can recall which profile he is on?