NBP part II

Marlow

joe_99 wrote: »

Eirs APQ file homes passed. Then ComReg's FTTH active customers (minus SIRO 50k customers).

SIRO is at something like over 80k customers atm. Also FTTH active customers figure from Comreg will also include Magnet, eNet and a few regional providers, even though that probably is not a large number.

If you have access to OpenEIRs APQ, then you also have access to the CLI list. And that will tell you, that it is 18% not having excluded multiple lines to premises, based on first week December data.

/M

KOR101

Marlow wrote: »

SIRO is at something like over 80k customers atm.

They've passed 290k homes, so that's 28% for SIRO.

Marlow

KOR101 wrote: »

They've passed 290k homes, so that's 28% for SIRO.

Couldn't find the last passed figure. Thanks for that. You could argue urban vs. rural, but SIRO is much more effective at marketing. That's for sure.

And Vodafone is pushing SIRO massively.

/M

joe_99

Marlow wrote: »

SIRO is at something like over 80k customers
/M

https://siro.ie/news-and-insights/gimme-fibre-celebrating-the-contribution-of-ftth-to-building-a-more-sustainable-future/

50k according to SIRO

joe_99

I estimate eir have around 95k active FTTH

Marlow

joe_99 wrote: »

https://siro.ie/news-and-insights/gimme-fibre-celebrating-the-contribution-of-ftth-to-building-a-more-sustainable-future/

50k according to SIRO

I'd say somebody has given her the wrong figures there. I'd be very worried, if out of over 84k orders only 50k had materialised.

But to get correct figures for OpenEIR, you need to look at OpenEIR data only. And it's 18% at most based on that.

/M

joe_99

Marlow wrote: »

I'd say somebody has given her the wrong figures there. But to get correct figures, you need to look at OpenEIR data only. And it's 18% at most based on that.

/M

It's a SIRO press release surely it is a simple number to check

wexfordman2

Emmanuel Lemon Louse wrote: »

The first 35 locations to be built have been announced. It seems they have sourced getter quality optics than open eir as the distance limit has been increased to 30km apparently.

https://www.irishtimes.com/news/ireland/irish-news/broadband-plan-first-35-areas-to-get-high-speed-service-are-revealed-1.4107462

There's a huge focus on the MAN towns in the first 35 locations. Practically all of them are enet locations.

That's interesting, but what do they mean exactly, what is the coverage area of each of these locations. For example, do they plan on finishing these areas completely, ie have all houses within 30km fibre range covered, or is it just a small localised deployment ?

Does anyone know if any maps are available for these areas?

tsue921i8wljb3

wexfordman2 wrote: »

That's interesting, but what do they mean exactly, what is the coverage area of each of these locations. For example, do they plan on finishing these areas completely, ie have all houses within 30km fibre range covered, or is it just a small localised deployment ?

Does anyone know if any maps are available for these areas?

There would be no maps as they have not even started to survey yet. What can be said is that the list of OLT locations that I have has 87 total enet locations being used with a further 118 open eir exchanges (plus 22 undertermined on islands) and of the first 35 locations 34, I believe, are enet with the only open eir being Kilkishen in Clare. Coincidentally, guess who has a home in Newmarket-on-Fergus which seems to be in the catchment area of Kilkishen? Our old friend Mr McCourt.

The 30km could work two ways. Either they will try to get as many premises passed by extending 30km from each of these 35 locations, which I thinks is unlikely myself, or they will merely start building in these locations extending several km then start on more of the 87 enet or move to some open eir. Either way it looks like this is the plan to take us up to 2022 so unless you are close enough to one of the 35 locations I would not expect service before then.

Why there is a reluctance on using open eir buildings in the first phase I'm not sure. Perhaps eir are looking for massive rent on using their buildings so it makes more sense for them to use their partner's locations.

Again the 300k and the refusal to use open eir fibre is being felt now. How much time and money is going to be wasted passing by premises that already have service in this plan.

Airwire: MartinL

wexfordman2 wrote: »

That's interesting, but what do they mean exactly, what is the coverage area of each of these locations. For example, do they plan on finishing these areas completely, ie have all houses within 30km fibre range covered, or is it just a small localised deployment ?

Does anyone know if any maps are available for these areas?

At this point of time, they have not even completed their surveys. The only map that is available, is the NBP Coverage Map at the DCCAE (https://www.dccae.gov.ie/en-ie/communications/topics/Broadband/national-broadband-plan/high-speed-broadband-map/Pages/Interactive-Map.aspx ) and that is a bit out of date.

- A basic deployment plan is expected Q1/2020.
- The first BCPs will be connected in Q2/2020.
- The infrastructure for partnering ISPs to connect to won't be in place until Q3/2020
- The pilot in Q4/2020 will see 10k premises passed and all BCPs connected.
- 2021 will see 130k premises passed with premises in every county.

That is as much information as is currently available.

Airwire: MartinL

Emmanuel Lemon Louse wrote: »

Why there is a reluctance on using open eir buildings in the first phase I'm not sure. Perhaps eir are looking for massive rent on using their buildings so it makes more sense for them to use their partner's locations.

OpenEIR generally does not want other providers equipment in their premises and with NBI, they would not only house NBIs equipment, but also the gear of providers, who direct connect to NBI opposed to Bitstream.

Nevermind the cost and the hassle to even get into the building, when work has to be done.

eNet on the other side is less complicated on that particular part of their portfolio.

tsue921i8wljb3

Airwire: MartinL wrote: »

OpenEIR generally does not want other providers equipment in their premises and with NBI, they would not only house NBIs equipment, but also the gear of providers, who direct connect to NBI opposed to Bitstream. Something that a good few providers would want to do, to ensure quality and control runaway cost from bandwidth charges.

Nevermind the cost and the hassle to even get into the building, when work has to be done.

eNet on the other side is less complicated on that particular part of their portfolio.

There doesn't have to be an interconnect at every open eir exchange though. They could have started building in more rural exchange areas and interconnected at larger open eir exchanges or alternatively at enet locations. Instead they have chosen to start by running yet more fibre through large regional towns.

Marlow

Emmanuel Lemon Louse wrote: »

There doesn't have to be an interconnect at every open eir exchange though. They could have started building in more rural exchange areas and interconnected at larger open eir exchanges or alternatively at enet locations. Instead they have chosen to start by running yet more fibre through large regional towns.

The department is paying for the stretch from their interconnects/PoP locations to the point where they meet the first houses.

It is part of the contract that the department signed and agreed to. It is also the reason, why no premises outside of the NBP ever will be connected to this network.

The elephant in the room here is: Why did the department not challenge this, before signing the contract.

/M

tsue921i8wljb3

Marlow wrote: »

The department is paying for the stretch from NBIs interconnects/PoP locations to the point where they meet the first houses.

It is part of the contract that the department signed and agreed to. It is also the reason, why no premises outside of the NBP ever will be connected to this network.

The elephant in the room here is: Why did the department not challenge this, before signing the contract ?

For NBI it is most likely the solution, where they have the least amount of outlay on their own side of things.

The other reason could be, that eNets infrastructure .. at least a lot of it .. is government owned. OpenEIRs infrastructure is not. Political decision to show the MANs are being utilized ?

/M

It probably is due to the costs in that using the enet locations is cheaper but then you recall articles like this from July 2019 and specifically this part:

One source said the official (Fergal Mulligan) gave the impression that not only would NBI be permitted to sell outside the intervention area but that it intended to do so.

Marlow

Emmanuel Lemon Louse wrote: »

It probably is due to the costs in that using the enet locations is cheaper but then you recall articles like this from July 2019 and specifically this part:

One source said the official (Fergal Mulligan) gave the impression that not only would NBI be permitted to sell outside the intervention area but that it intended to do so.

Well, that question was asked on Thursday and even if it was considered in July, it certainly is a big no no now, after the contract is signed.

The answer by NBI was simply along these lines: "If we sell outside of the intervention area, we will have to return the build cost for the infrastructure between the intervention area and the interconnect back to the department. Only premises within the intervention areas will be served." They won't do white label or retail either. Wholesale only.

There is a provision to bring connectivity to masts. There is a cost involved though. And it was indicated, that that connectivity also only may be used to service premises, that are in the intervention area. I believe there are still questions to be resolved with the department there.

There is also the case, where a premise with one Eircode orders more than one connection: It is possible, but only the first connection to said premise will be covered by the NBP funds. For the second connection the full connection costs will be charged, which in some cases could be 5k+.

All of this was made very clear in the presentation last Thursday.

/M

heavydawson

30km radius overlay map for all points references in the Irish Times article

SkepticQuark

heavydawson wrote: »

30km radius overlay map for all points references in the Irish Times article

So given that covers most of the country, probably not doing that all in 2021. I'd imagine they're talking installation of points and beginning of rollouts?

tsue921i8wljb3

heavydawson wrote: »

30km radius overlay map for all points references in the Irish Times article

I think it's highly unlikely they'll go 30km from each OLT. They have 205 mainland OLT locations listed so it's more probable that the individual clusters will be much smaller.

Marlow

Emmanuel Lemon Louse wrote: »

I think it's highly unlikely they'll go 30km from each OLT. They have 205 mainland OLT locations listed so it's more probable that the individual clusters will be much smaller.

They have to. Every spur takes away from the 30km. Assume an average of 150m for an access spur. 32 connections per cluster (that is a confirmed figure). That is 4.8km in spurs.

Leaves you 25km'ish for the cluster run. And that isn't going to be in a straight line either.

20km radius or less is probably a safer bet.

/M

tsue921i8wljb3

Marlow wrote: »

They have to. Every spur takes away from the 30km. Assume an average of 150m for an access spur. 32 connections per cluster (that is a confirmed figure). That is 4.8km in spurs.

Leaves you 25km'ish for the cluster run. And that isn't going to be in a straight line either.

20km radius or less is probably a safer bet.

/M

I have no idea what any of that means bar they are using a 1:32 split ratio.

tsue921i8wljb3

Actually I think I understand what you're trying to say. You're multiplying 32 x 150m and taking this out of 30km. That's not how it works. Each point can be a maximum of 30km from the OLT. The cumulative total does not matter.

Edit: The cumulative bit is strictly not true. 32 points at 30km each would be a cumulative max of 960km (32 x 30)

Marlow

No. Each port of the OLT can reach 30km. But the 32 premises are per port.

That means the complete fiber length hanging off each individual port can not exceed 30km. Anything from the OLT to the ONT is passive. There is no amplification.

The fibre runs from the port to DP. This may happen in multiple phases. The length of the furthest premise away can not exceed 30km including spur.

Now, I agree, that in theory spurs along the route may not take away from the length. But some safety margin will also have to be build in for signal loss from splice and resplice if the fibre ever breaks.

That and the fact, that the fibre does not run in a straight line.

So they will never achieve a 30km radius from the OLT. Hence 20 km is a safe bet. Maybe 25km.

/M

tsue921i8wljb3

Marlow wrote: »

No. Each port of the OLT can reach 30km. But the 32 premises are per port.

That means the complete fiber length hanging off each individual port can not exceed 30km. Anything from the OLT to the ONT is passive. There is no amplification.

/M

Marlow, I was trying to be tactful in my last reply. But you had to go and double down. You clearly don't have the first idea how any of this works. You started a couple of years ago stating that each DP had 2.5Gb/s and your knowledge seemingly has not improved. You're way off on this again.

Is there anyone else reading this who can confirm that I'm right?

Just to clarify that Marlow is stating that the total distance used by cable in a 1:32 split is 30km (assuming 30km optics).

I'm saying that each point in a 1:32 split can be max 30km (again with 30km optics) from the OLT (however unlikely) and assuming the optical budget is met so the cumulative distance of cable would be 960km.

Marlow

Emmanuel Lemon Louse wrote: »

I'm saying that each point in a 1:32 split can be max 30km (again with 30km optics) from the OLT (however unlikely) and assuming the optical budget is met so the cumulative distance of cable would be 960km.

That can only happen, if the split happens at the OLT. But that is not how the cluster is build in real life.

If the DP is 5 km from the port, then each premise can only be 25km from there.

It is a matter of where the split happens. The end to end stretch from OLT to ONT can not exceed 30 km.

/M

Marlow

Anyhow .. fact is:

- Maximum distance between optical interfaces (any): 30km
- the fibre is not ran in a straight line

Bottom line: you never achieve 30km radius. And that IS the point i was making.

/M

tsue921i8wljb3

Marlow wrote: »

That can only happen, if the split happens at the OLT. But that is not how the cluster is build in real life.

If the DP is 5 km from the port, then each premise can only be 25km from there.

It is a matter of where the split happens. The end to end stretch from OLT to ONT can not exceed 30 km.

/M

You have no clue what you are talking about. No clue.

You're backtracking as you usually do now that you've been found out. Trying to move the goalposts. You constantly post absolute crap on here and get away with it.

You might want to read this page and in particular pay attention to the planning with splitters section. You'll notice that the total cable distance on the GPON port in the image is 42km on a GPON port that you claimed could have a cumulative total of 20km. And that's with a part 1:64 split which weakens the optical signal further.

https://help.ubnt.com/hc/en-us/articles/115011654907--UFiber-GPON-Designing-a-GPON-Network

All that matters is that each individual ONT is within the optical limits.

Marlow

Marlow wrote: »

Now, I agree, that in theory spurs along the route may not take away from the length.

Emmanuel Lemon Louse wrote: »

You're backtracking as you usually do now that you've been found out. Trying to move the goalposts. You constantly post absolute crap on here and get away with it.

So you completely omitted this .. where I admit, that I can see what you're saying and start slating me instead. Thanks.

Marlow wrote: »

Anyhow .. fact is:

- Maximum distance between optical interfaces (any): 30km
- the fibre is not ran in a straight line

Bottom line: you never achieve 30km radius. And that IS the point i was making.

And you don't agree with this ?

/M

tsue921i8wljb3

You said this originally

They have to. Every spur takes away from the 30km. Assume an average of 150m for an access spur. 32 connections per cluster (that is a confirmed figure). That is 4.8km in spurs.

Leaves you 25km'ish for the cluster run.

This implies that you believed that 32 x 150m contributed 4.8km which you then subtracted from 30km as if it was some magical total that had to be observed. Why else would you have subtracted it from 30km.

This makes no sense and shows that you don't understand how GPON networks work.

I then tried to correct you by stating that each ONT could be 30km from the OLT leading to a much greater cumulative cable distance than 30km.

Again you replied

No. Each port of the OLT can reach 30km. But the 32 premises are per port. That means the complete fiber length hanging off each individual port can not exceed 30km.

Again, you seem to be saying that the cumulative total of cable on a port can only be 30km which is wrong and shows a misunderstanding of the technology.

Now that I have pointed out that you don't know what you are talking about you are waffling and spoofing, like you usually do, trying to say you didn't mean what you actually said.

Also please don't try to lecture me about how clusters are built in "real life". I deal with this stuff on a daily basis and I don't need some amateur giving me lessons.

To be honest I have wasted enough time with you so I'll leave it up to others to see who they believe, if they even care, which they probably don't.

babi-hrse

Jesus this has exploded
I value Marlow input as he's coming from a point in the network and much of what he has said is true which has added alot of perspective into what is otherwise a repetitive task
He has been wrong on a couple of points but I believe everything he has said on procedure. he was wrong on distance on VDSL in that it was useless after a point much shorter than it really was but this is conjecture as different distances are possible further than he specified but this depends on the condition of the cable but by and large he is right on alot of topics.
As for ftth nearly no lines ftth run more than 9km from exchange assuming the aggregate node at a small exchange facing its ok parent exchange subtract from the signal loss in a major way
I have not seen a dp that was worse than -14db @1550nm I have been informed that the sc\apc connections are a loss of 0.3 I don't presume to know the main network cable splices that are fused but I have it on good authority that the losses are negligible as their work gets checked and if it fails strict standards it has to be rejoined.
This is lads saying they don't get paid unless it passes.
But I would be interested in knowing the loss if another splitter was added as there area few DPS that are full and I would like to know if it's possible it can be expanded to accommodate more

tsue921i8wljb3

babi-hrse wrote: »

Jesus this has exploded
I value Marlow input as he's coming from a point in the network and much of what he has said is true which has added alot of perspective into what is otherwise a repetitive task
He has been wrong on a couple of points but I believe everything he has said on procedure. he was wrong on distance on VDSL in that it was useless after a point much shorter than it really was but this is conjecture as different distances are possible further than he specified but this depends on the condition of the cable but by and large he is right on alot of topics.
As for ftth nearly no lines ftth run more than 9km from exchange assuming the aggregate node at a small exchange facing its ok parent exchange subtract from the signal loss in a major way
I have not seen a dp that was worse than -14db @1550nm I have been informed that the sc\apc connections are a loss of 0.3 I don't presume to know the main network cable splices that are fused but I have it on good authority that the losses are negligible as their work gets checked and if it fails strict standards it has to be rejoined.
This is lads saying they don't get paid unless it passes.
But I would be interested in knowing the loss if another splitter was added as there area few DPS that are full and I would like to know if it's possible it can be expanded to accommodate more

And who explained to you why you hadn't seen a lower value? Was it Marlow?

https://www.boards.ie/vbulletin/showpost.php?p=108600910&postcount=5302

It's dBm by the way. A decibel (dB) is a dimensionless unit.