Stats, versions etc help

Okay, finally hooked up a machine to LAN2 and got DSLstats running.

Is there an idiots guide somewhere to what all the stats actually mean?

One thing I was curious about is interleaving, given that our speed has gone down so much since we got it. Apparently it's on at "823", but I don't know if that's good bad or indifferent since it could be 823 furlongs for all I know at the moment!

I noticed it still has the web interface, which I think means it's still running the old firmware. It's been up for 20 days. Is it worth a reboot, will that get the new firmware etc? (I never dare touch it in case the DLM thinks there's something wrong with the line and lowers it even further)

Other main stats are up 16188 @ 5.1dB down 51983 @ 5.6dB fwiw

I have a reasonable knowledge, and have explained the stats in a few posts in the past, but don't have time to do that so much these days.

If you search for my posts here, you'll see me talk about interleaving, FEC, INP, delay in more detail.

In particular, you can see that the R and N parameters will show you the overhead of FEC (R/N as a percentage usually amounts to around 20%). The D and I (capital i) figures tell you the scale of interleaving in terms of bytes buffered (D * I), but the total delay, in ms, comes from the "delay" value.

INP and delay values are the figures set by DLM when it asks for interleaving to be turned on. An INP value of 3 is the usual initial setting once DLM intervenes (INP = impulse noise protection, and defines how big a burst of noise must be protected against). A "delay" value of 8 is the first setting there, meaning 8ms. Larger values in either of these means DLM thinks you have bigger problems.

One day, we'll hopefully switch to PhyR, and have no need of FEC or interleaving.

Many thanks, every little helps. I appreciate people don't have time to explain it over and over I've been having a search around and it's tricky finding the search terms that turn up good results, since the various forums out there have "signal to noise ratio" problems of their own!

I have just found mention of R and N elsewhere as well, and calculated ours at 20% as you say. It makes perfect sense as with a "first day or two" speed of ~60mbits it explains why it's reading 47.5 now, the maths add up.

It's far from the end of the world, it's just that when BT's checkers etc say we should get 60-odd it's very frustrating, it's like a dangling carrot. If there's anything I can do to reduce the errors and therefore get the DLM to relent, I'll be a happy bunny. I may resort to wandering around the house with a MW radio set to 800khz looking for noisy PSUs as one person somewhere on the internet came up with. I am seeing a rather high number next to FEC, so unless that drops I can't see it removing it.

PHYR sounds promising if there's any indication they're considering it. I look forward to vectoring as well (not sure, that may be rolling out now?) as that should help curb any "meh, it's just crosstalk" hand-waving.

Pretty much anything that will make the Dynamic Line Nazi chill out a bit is all good.

Edited by deleted (Mon 02-Dec-13 22:56:56)

In reply to a post by LakesGeek:

PHYR sounds promising if there's any indication they're considering it. I look forward to vectoring as well (not sure, that may be rolling out now?) as that should help curb any "meh, it's just crosstalk" hand-waving.

The requirements for CP-provided modems (or modem/routers) as an alternative to the Openreach modem demand a few things of note:
- Profile 17a
- Vectoring capability
- PhyR capability
- SRA capability

They suggest that BT won't be bothering with profile 30a, and that they are intending to at least consider the other items.

Vectoring is in a public trial phase, with 4 or 5 cabinets having been activated, and real users in the trial (every user on the cabinet, with all of them at 80Mbps).