Print

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

[Chrysalis]

Ok so as many of us know banding seems to be (almost) permanent.

Today it got removed on my line.

The only thing I can observe is my crosstalk reduced a lot about 5am 5 days ago, and it didnt come back during that time, so a theory is that banding may need a lot of spare snrm for it to be removed, my DS snrm was over 11db during those 5 days.

[j0hn]

Did the missing crosstalker help reduce ES numbers at all (or FEC if interleaved)?

[Chrysalis]

nope is same, ES also seems same at the higher sync speed as well.

My US ES increased a couple of weeks back for no obvious reason, so I guess ES isnt always explainable.

[underzone]

Wow! Nice sync speed. Probably worth all your recent disruption.

PS. Which firmware are you running on your 8800NL?

[Chrysalis]

underzone 2.32e.dh4

and yeah, a 3 week outage for the better tie pair, yes thank you.

[Chrysalis]

can see after some uptime, DS ES is about 1.5x higher with the lower snrm, and US ES is reverted to under 0.5 ES/hour again.

[kitz]

Nice work. The cab rebuild really did you a favour and improving your SNRm 
 
Can you recall what caused it to get banded at 74Mbps? I can see a large spike of Err/Secs on Apr 12th after the SNRm had been running at 2.9dB for a while, but can't see if there were a lot of retrains.  Banding seems a bit harsh for what on MDWS looks like a single bad day or Err/Secs.

The reason Im curious is that I've had much higher Err/Secs than you have and the only diff being that I can see from MDWS is that your SNRM was low.   
You may have a point about needing a lot of spare SNRM, and why some suggest self capping.  On face value it certainly seems to have been what did so on your line.   

[Chrysalis]

My theory was that DLM was able to put 2 pieces of data together to decide to band the line, I think you disagreed with me before but here is the theory.

1 - The line required stability action after hitting red for 24 hours. (high ES)
2 - The line had synced high during a wide area event (power cut), and the red 24 hours was during this high sync.

It seems if this happens it will band instead of interleave.