Print

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

[roseway]

Yes, it's my baby.

[roseway]

As we're not entirely sure what the router is reporting, I'm going to offer two alternative ways of dealing with the situation of the reported CRC total in a sample being less than in the previous sample:

1. Zero values will be plotted instead of negative values (the default option)
2. Negative values will be converted to positive values

There will be an extra option on the Graphs configuration page to make this choice.

Eric,

I can't give you conclusive proof of this, other than the obvious, but I believe that I saw a similar issue with the Bitswap/min graph having -ve values on the y axis in 2.7.3.  In 2.8, they are flat-lined zero.  The Bitswap/tone graph, and common sense, suggest that it's (almost) never zero, so is this another example of receiving bad data on the sampling intervals?   Cetainly the bitswap (total) counts from xdslcmd are not zero.

It's not a big deal at all, more of a minor inconvenience.  And sorry to be a party-pooper, but I'm using the RSW version. 

Sorry Colin, I seem to have ignored your message, but it wasn't intentional. I'll be looking at this today.

[snadge]

Yes, it's my baby.

great! when did you start making this? are you always working to get it operational with other routers? it works great on my 834N, i particularly like the CRC average counts (per hour etc) as Iam monitoring those on my line at moment

[roseway]

great! when did you start making this? are you always working to get it operational with other routers? it works great on my 834N, i particularly like the CRC average counts (per hour etc) as Iam monitoring those on my line at moment

I started last autumn. If you're interested in the details, it's all recorded in the file CHANGELOG which is in the zip file you downloaded.

I add extra routers to the compatibility list from time to time, but I don't intend to go beyond those which have Broadcom chipsets. The next version will include some Thomson/Technicolor models which weren't previously supported.

[snadge]

hey its great..well done Eric, just wished i'd known about it sooner...  I have one idea that might be hard to implement..but i seen it on mrmt32's Sky Router Tool (Classic Version) - you know the SNR graph sits behind the bitloading? well it would be great if the SNR was proportional to the Bit-loading graph in terms of how much SNR the BL is using.. I see it sits tall above it.

see below, Ive tried to feebly draw in what I mean (where the SNR might go to) , on my stats I have a 5db Noise Margin, so (if what Iam getting at is right) the average margin per tone is 5db, in the image below the SNR is in 5db increments so ive drawn a black line representing the 5db between the SNR and Bit-loading... is that right?  just I seen something similar on Sky Router Tool in that it looked proportional - no biggy, just an obvservation. EDIT: have added screenshot of SRT below - used vs actual SNR Graph

....shoot me down in flames now hehehe

[roseway]

The scale for SNR is on the right hand side of the graph. The way that it 'sits tall' is just the result of the scale I chose (deliberately) so as to make it stand clear of the bitloading data. You can't draw any conclusions from the gap between the SNR and the bitloading because they're different units (dB and Bits respectively). It's just an overlay of one graph on another.

But I may have misunderstood what you're asking for.

[Edit] I've just seen the extra two graphs you included. Each bit of data carried by a tone requires 3 dB of SNR, so I imagine that 'Max loading' is the SNR value divided by three. I could add that to the graph, although I'd need to think about it a bit, because I don't want to make it too cluttered.

[snadge]

I dont mean 'Max Bit-loading' (although that seems clever too) , I refer to the bottom one in green, with the 'used' SNR shown just like the bit-loading graph and it sits over the 'actual' SNR - it sort of gives you a visual perspective of what your 'using' in terms of SNR... and the difference between the two is whats used as 'margin' ..am i right in saying that? taking on board you saying each bit requires 3db of SNR, well if you look at your graph.. each 5db lines up with 2bits, so one woould simply amend the axis so instead of the SNR reading 5, 10, 15, 20 - it would be 6, 12, 18, 24 etc because every 2 bits equal to 6db... that way the bit-loading graph would be "proportionate" too the SNR graph, you can literrally see how much SNR the bits are using up - if that makes any sense? like I said before a sort of visual perspective of how much signal your bandwidth is using and how much is margin (in correct 'proportions'...christ if i use that word again!! hehe, sorry my vocabulary int great) - looking back at mrmt32's graphs he has done exactly that - the SNR goes up to 60db and bitloading 20 bits which works out at 3db per 1bit

It was just something I thought I would point out - either way the program functions perfectly without it doing it like that.. I just thought that graph..(or rather how it was done) on Sky Router Tool was a nice touch... it shows how much SNR is used by popping in the Bit-Loading graph over the top in proportion to the amount of SNR its using...if thats right lol 

if he was really clever he could of put all 3 graphs together... used SNR (current bit-loading), MAX Bitloading, SNR ..in that order

edit: can I ask how the CRC averages are calculated? they dont seem to be working out from 'last sync' (uptime) - currently i have uptime of 20.5 hours ands 54 CRC's = 2.63 CRC's per hour - but RS-W says 4.05?

EDIT 2: i see its 24 hours , so its estimating it when its running less than 24 hours?...is that right? is there anyway for a user to change it to show it since UPTIME (last re-sync) as other tools like SRT show it from system uptime and total CRC's - I just want CRC average from last re-sync as Iam testing error counts per SNR Margin settings...at moment I am using calculator lol

[snadge]

sorry...ive made a lot of changes to the above post... just in case you have read it before the edits Eric, I think its more clear on what I mean, simply amending the SNR Axis to a max of 60db would do it as you already have max of 20 for bit-loading.. this would yield the 3db per 1 bit and keep the graphs in scale with each other.

[roseway]

OK, I see your point now, and I'll have a think about rescaling the SNR graph to equate with 3 dB per bit.

Concerning your other point, the average error rates are calculated on periods of up to 24 hours. The start of the period currently being viewed is shown above the table of data. The values in the table are the actual number of errors in the period divided by the actual length of the period, converted to per-second, per-minute, etc. values. When you've been running without a resync for more than 24 hours you can look back to earlier 24 hour periods which have been completed.

I realise that this isn't exactly what you're looking for, and I'll have a think about incorporating values since the last resync, but unfortunately not all routers accurately report the DSL uptime. So unless the program was actually running at resync time so it could record the event, it may not be possible in all cases to get an accurate set of figures.

[ryant704]

Started crashing for me...

When I go on FEC Errors and Logarithmic is enabled it will say it's crashed. Clipping only allows me up to 4995, it was fine for days then last night it's started crashing every time I go on it and it is enabled.

Problem signature:
  Problem Event Name:   APPCRASH
  Application Name:   rsw.exe
  Application Version:   2.8.0.0
  Application Timestamp:   00000000
  Fault Module Name:   StackHash_c9e3
  Fault Module Version:   0.0.0.0
  Fault Module Timestamp:   00000000
  Exception Code:   c00000fd
  Exception Offset:   73f5e254
  OS Version:   6.1.7600.2.0.0.256.1
  Locale ID:   2057
  Additional Information 1:   c9e3
  Additional Information 2:   c9e3143907df77b8e420fb642424b741
  Additional Information 3:   a56e
  Additional Information 4:   a56e55d65c7529f54c14f924428fec56

[roseway]

Started crashing for me...

When I go on FEC Errors and Logarithmic is enabled it will say it's crashed. Clipping only allows me up to 4995, it was fine for days then last night it's started crashing every time I go on it and it is enabled.

This sounds like the same issue as a couple of other people have reported (the crash is due to trying to take the log of a negative number) and I believe it's fixed in the next version.

[ColinS]

Sorry Colin, I seem to have ignored your message, but it wasn't intentional. I'll be looking at this today.

Not at all.  So many things to do, so little time.  I know. 

[roseway]

Not at all.  So many things to do, so little time.  I know. 

I've made some changes which will hopefully deal with your bitswaps issue.

[ColinS]

Not at all.  So many things to do, so little time.  I know. 

I've made some changes which will hopefully deal with your bitswaps issue.

Thanks.