Print

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

[Chrysalis]

thanks BS is indeed useful info, so e.g. when an engineer visits we can ask him to show us this data or if he has checked it.

Any idea if the threshold for red is different to DLM thresholds?

[Chrysalis]

I think my concern over this is just the problem where the end user may be getting well over the average UK speed BUT prior to the call they were getting even more above average and suddenly not. In my view comparisons to UK average speeds is not correct, a sudden drop of a significant amount must be investigated irrespective of what they now get compared to UK average speed. Just saying 'you're above average even now' is simply not acceptable.

Stuart

This is correct which raises another question, will GEA failures still count as a failed test?

e.g. if >25% of sync speed is lost in a short time period a line will auto fail the GEA test "REGARDLESS" of what the sync speed and error rate is.

My guess is for a line to be RED it is probably going to have to be in quite a bad state as its probably set very conservatively to keep costs down for openreach.  I also wonder if the RED/GREEN is just a quick way for engineers to check GEA status, so GREEN means GEA is passed?

[Chrysalis]

To be serious this new line health indicator would need previous history and current history from the end-users line i am not sure the DLM database holds so much data for each line line in the UK there has to be what's called a data limit for each end-users line.

I guess it could hold upto 3 to 5 years worth of stats for each end-users but it won't look like Roseways stats or Bald_Eagle1 stats it will be BT OR stats

Your idea makes too much sense.

I thank BS for the information is much appreciated but like most changes to openreach SFI conditions, it is designed as a way to stop CP's and end users pushing for openreach to carry out work on a line.

e.g. my belief has always been openreach should "prove" a drop in speed is due to crosstalk rather than "assuming" it is, and some may remember I posted a while back there needs to be some kind of service which guarantees all physical aspects of the line get checked on a visit regardless if a JDSU passes or not, but of course openreach dont want to pay engineers to spend hours on each callout (and probably also CPs want to pay openreach as little as possible) so this is the situation we have.  I would expect it be easy to cross reference the history of a line and check the installation database to see if the time matches any new installations in the area if yes, then is evidence its crosstalk caused, if no, start an investigation.

So I see both sides, it must be frustrating to deal with people who e.g. have triggered a call out for a loss of say 2% of sync speed, but it works both ways it be frustrating for customers to be lied to by engineers and their isp to brush off faults. e.g. I had an engineer tell me the BT wholesale supplied speed estimate was complete rubbish and that speed was for those next to the cabinet (so he claimed the max possible sync speed on my cabinet is 72mbit at 0m distance with no crosstalk), he was been a bit aggressive so I didnt push it with him.  Shame not all engineers are not like my install engineer, but even the better engineers will get in trouble if they do work when they not supposed to, as my engineer backs up what BS has said that if tests pass then they supposed to walk.

[Black Sheep]

But that wouldn't help if your line, wiring or environment has always been bad, so that your low speed is the same low speed you were getting in the past.

Isn't a boost visit some kind of special visit, always paid for by the CP, aimed at improving the speed, rather than fixing some fault? Surely there's no point in doing a boost visit then, if the indicator is already green.

As pointed out previously, the speed will be compared against the national average for a given attenuation ..... so it will help. 

The only difference with a 'Boost' visit and its sister 'SFI' visit, is that we can do a Hub swap if the original is proven faulty. Nothing special.

[Black Sheep]

thanks BS is indeed useful info, so e.g. when an engineer visits we can ask him to show us this data or if he has checked it.

Any idea if the threshold for red is different to DLM thresholds?

Chrys, we won't be privy to the data .............. we will get a 'Green' or a 'Red' indicator. IF all tests pass at the EU's premises AND we have a 'Green' indicator on the task. Then we close the job off. No network intervention will be done, that is all it is. 

[Black Sheep]

Chrys ............... "some may remember I posted a while back there needs to be some kind of service which guarantees all physical aspects of the line get checked on a visit regardless if a JDSU passes or not, but of course openreach dont want to pay engineers to spend hours on each callout (and probably also CPs want to pay openreach as little as possible)"

A lovely thought, but business and financial suicide. It simply is not plausible.

The problem with forum posting is one tends to think of just their situation ..... can you imagine checking all physical aspects on a 200+ carrier pole installation, never mind the UG cable runs ?? We have just one like this on my patch, but in Scotland and Wales, it is common-place.

We have to provide a 'one-glove-fits-all' service, and one that has been agreed with the CP's and Ofcom. We have .... it is SIN 349. Coupled with 'DSL Close out tests' and 'Eclipse CIDT' tests ...... the vast majority of faults will be picked up via one of these tests.

I've said it before on here before, if you could see the actual information our tech-heads use and work with, your mind would be blown. The info on DLM alone, would have B*Cat, Kitz, 4Candles and JGO asking to leave the building.  Openreach are part of a FTSE100 company, they know what they are doing ......... and I'm sure they appreciate any feedback on where EU's think they've got it wrong. 

[kitz]

I wonder if this is anything to do with the new  BRAT (Brandenburg Rapid Analysis Tool) that BTw is/has released this year available as part of KBD. 

I mentioned in another thread a several weeks ago that Ive been trying to get Plusnet to run one on my line, but Ive been told they cant run it because I have an open fault.   The 'fault' I have open isnt really a line fault, but more to do with the fact that my line is one of those being monitored because the peak time slowness can be replicated.   I know they did try to run it because it actually knocked me offline when it was attempted to run.

RRT has been available for a while for adsl2+ lines, although this data is collected by the element managers its not part of the DLM and information is stored in a separate RRT database. 
The new BRAT version supposedly contains more depth analysis and is for NGA lines (FTTC)

This is an example of an adsl2+ line showing hourly sync speed and Err Secs, but it can be drilled down to view 15min slots.   


If BT are storing this data and making it available to the SPs, then it makes sense that some of or the same data could be used by the OR engineers?

[Black Sheep]

I've personally had access to BRAT for a few years now. Alas, only on ADSL profiles ...... it is not available on my GEA profile. 

I would hazard a guess you are correct with regard to BRAT being used on BT classic circuits, but we have to rely upon the CP's own version of BRAT when looking at DSL fault history, (We request that they send over a data-pack).

BRAT has gotten better with age, it used to summarise all its findings as REIN in a lot of previous cases, but when the attenuation graph was perused it was obvious a HR was allowing in unwanted 'Noise' that BRAT hadn't picked up on.
Now though, it does for the most part recognise if there is a potential 'HR' or similar network issue, (wet joint etc).

You can never, ever achieve a better test than a man in a van turning up at your door and running thorough tests, but I have to admit that the graphing data available to us, is far better of late.     

[loonylion]

You can never, ever achieve a better test than a man in a van turning up at your door and running thorough tests, but I have to admit that the graphing data available to us, is far better of late. 

depends if it's a battle hardened BTOR man, a padawan 6 weeks out of college, or a kelly's cowboy

[kitz]

I've personally had access to BRAT for a few years now. Alas, only on ADSL profiles ...... it is not available on my GEA profile. 

I would hazard a guess you are correct with regard to BRAT being used on BT classic circuits, but we have to rely upon the CP's own version of BRAT when looking at DSL fault history, (We request that they send over a data-pack).

BRAT has gotten better with age, it used to summarise all its findings as REIN in a lot of previous cases, but when the attenuation graph was perused it was obvious a HR was allowing in unwanted 'Noise' that BRAT hadn't picked up on.
Now though, it does for the most part recognise if there is a potential 'HR' or similar network issue, (wet joint etc).

You can never, ever achieve a better test than a man in a van turning up at your door and running thorough tests, but I have to admit that the graphing data available to us, is far better of late. 

There was supposed to be some big changes to BRAT starting Jan this year, including bringing in a more in-depth Brandenburg test for NGA which is why I was asking Plusnet to try do one on my line, so I could see what the new features look like when run on a live line.

For others that dont have access to BS's info, below is a screen cap of information provided by the 21CN version..  which I should imagine is what BS means about how it has improved.   The aim is to bring this over to GEA-FTTC under the name of  "Brandenburg NGA" and is supposed to be an enhancement of the existing NGA tests..  specifically to assist in the cause of a drop in speed of the circuit and to help pinpoint interference problems.

It was mentioned in their roadmap that they have plans for some change in Sep 2015, screen cap also below.. which is why I wondered if the line health indicator did have something to do with the new BRAT capability.
 

[Black Sheep]

You can never, ever achieve a better test than a man in a van turning up at your door and running thorough tests, but I have to admit that the graphing data available to us, is far better of late. 

depends if it's a battle hardened BTOR man, a padawan 6 weeks out of college, or a kelly's cowboy

  Cheers 

[Chrysalis]

I wonder if this is anything to do with the new  BRAT (Brandenburg Rapid Analysis Tool) that BTw is/has released this year a

I've personally had access to BRAT for a few years now. Alas, only on ADSL profiles ...... it is not available on my GEA profile. 

I would hazard a guess you are correct with regard to BRAT being used on BT classic circuits, but we have to rely upon the CP's own version of BRAT when looking at DSL fault history, (We request that they send over a data-pack).

BRAT has gotten better with age, it used to summarise all its findings as REIN in a lot of previous cases, but when the attenuation graph was perused it was obvious a HR was allowing in unwanted 'Noise' that BRAT hadn't picked up on.
Now though, it does for the most part recognise if there is a potential 'HR' or similar network issue, (wet joint etc).

You can never, ever achieve a better test than a man in a van turning up at your door and running thorough tests, but I have to admit that the graphing data available to us, is far better of late. 

There was supposed to be some big changes to BRAT starting Jan this year, including bringing in a more in-depth Brandenburg test for NGA which is why I was asking Plusnet to try do one on my line, so I could see what the new features look like when run on a live line.

For others that dont have access to BS's info, below is a screen cap of information provided by the 21CN version..  which I should imagine is what BS means about how it has improved.   The aim is to bring this over to GEA-FTTC under the name of  "Brandenburg NGA" and is supposed to be an enhancement of the existing NGA tests..  specifically to assist in the cause of a drop in speed of the circuit and to help pinpoint interference problems.

It was mentioned in their roadmap that they have plans for some change in Sep 2015, screen cap also below.. which is why I wondered if the line health indicator did have something to do with the new BRAT capability.
 

vailable as part of KBD. 

I mentioned in another thread a several weeks ago that Ive been trying to get Plusnet to run one on my line, but Ive been told they cant run it because I have an open fault.   The 'fault' I have open isnt really a line fault, but more to do with the fact that my line is one of those being monitored because the peak time slowness can be replicated.   I know they did try to run it because it actually knocked me offline when it was attempted to run.

RRT has been available for a while for adsl2+ lines, although this data is collected by the element managers its not part of the DLM and information is stored in a separate RRT database. 
The new BRAT version supposedly contains more depth analysis and is for NGA lines (FTTC)

This is an example of an adsl2+ line showing hourly sync speed and Err Secs, but it can be drilled down to view 15min slots.   


If BT are storing this data and making it available to the SPs, then it makes sense that some of or the same data could be used by the OR engineers?

kitz is this new info on your website for me to read? also is this info he posted on the mod section or something?

[kitz]

kitz is this new info on your website for me to read?

No its there on the DLM page.  I included a link to it in my post, but here it is again as a separate link
http://www.kitz.co.uk/adsl/DLM_system.htm#KBD

also is this info he posted on the mod section or something?

No not at all.  It was me who brought up KBD.
It was me who expanded on the topic purely because I was already aware of BRAT and the impending changes this year... and I was wondering it the two were related.  It was me who added the screenies which came from elsewhere simply because I thought others may not be aware of what is available from KBD.   Its been in place for many years for adsl2+, just not many ISPs seem to use it.

[Chrysalis]

yep I found that already, but not the new one about the upcoming GEA stuff.  I dont see that 2nd screenshot on the page.

[kitz]

I'm lost, Im not sure what you mean?
Click on the image it shows as long as you click in the right place.