IOL Contention Ratio Question

ZENER

If this has been covered already I'd be greatful if someone would point me to it, otherwise . . .

Shortly after Eircom reduced the wholesale price of BB to other ISP's I got a call from IOL asking if I'd like to avail of the new cheaper price or stay at the same subscription level (€48) and enjoy a higher 16GB cap and 20:1 contention ration.

I opted for the latter, now the 16GB cap is easy enough to verify and I have done but how can I tell if I'm benefiting from the lower contention ratio? My connections are always strong and never suffer slowdown but then they never really did even before this offer.

Can anyone shed any light on this?

ZEN

OfflerCrocGod

There is no real way for us to be certain that we are actually on a 24:1 ratio. We just afto trust what we are told

dub45

OfflerCrocGod wrote:

There is no real way for us to be certain that we are actually on a 24:1 ratio. We just afto trust what we are told

And further to that is there any proven benefit from being on the 24/1?

The recent 'non' problem which UTV had with their bandwidth seemded to affect all users who were on at the particular time irrespective of which contention there were on so I have been wondering is there any real point in paying the extra amount.

Moriarty

You get double the cap for a start. When uptake increases odds are a difference between the services will be evident. Or maybe not, who knows.

dub45

Moriarty wrote:

..............Or maybe not, who knows.

Well thats my question - does anybody actually know for definite? otherwise its 180 euros a year for what.............(apart from the cap which is a different issue obviously)

Ripwave

dub45 wrote:

Well thats my question - does anybody actually know for definite? otherwise its 180 euros a year for what.............(apart from the cap which is a different issue obviously)

Someone might interpret that as "If I'm paying €180 more than he is, I damn well want to know that he's getting a worse service than I am".

If the 48:1 users aren't experiencing contention at the exchange level, neither will the 24:1 users. On the other hand, 24:1 users can download twice as much as 48:1 users, and can therefore cause twice as much contention! (or cause the same contention for twice as long, if you prefer).

dub45

Ripwave wrote:

Someone might interpret that as "If I'm paying €180 more than he is, I damn well want to know that he's getting a worse service than I am".

If the 48:1 users aren't experiencing contention at the exchange level, neither will the 24:1 users. On the other hand, 24:1 users can download twice as much as 48:1 users, and can therefore cause twice as much contention! (or cause the same contention for twice as long, if you prefer).

That someone would be a very nasty type that you would never meet on boards.:) The proper interpretation is "I am paying €180 more - what am I getting for it?" The increased cap is completely separate issue - can anyone say for definite it is worth paying for the reduced contention rate?

As a previous poster seems to confirm we cannot even prove we are on the reduced rate and the recent UTV non-problem seemed to affect all customers who were on at a particular time equally so whats the point of it?

Has anyone who has changed back to the 48/1 noticed any difference? (apart from being €15 a month richer)

Moriarty

dub45 wrote:

The increased cap is completely separate issue - can anyone say for definite it is worth paying for the reduced contention rate?

It's completely seperate from a technical pov, but from deciding which package to get it certainly isn't.

dub45 wrote:

the recent UTV non-problem seemed to affect all customers who were on at a particular time equally so whats the point of it?

That had nothing to do with contention at the exchange, it was a problem that occoured by accident rather than by design. As more people sign up for DSL, a difference will probably become apparent.

dub45

Moriarty wrote:

It's completely seperate from a technical pov, but from deciding which package to get it certainly isn't.

I am not questioning the package just the contention ratio as an issue in itself.

That had nothing to do with contention at the exchange, it was a problem that occoured by accident rather than by design. As more people sign up for DSL, a difference will probably become apparent.

Accident? Thats too charitable a description (in my opinion!)
Design? Probably too Machiavellian

More likely somewhere in between:)


Probably become apparent - but when? Anyone know what decided Eircom to actually offer the product at this stage when there are so relatively few people still on bb?

Ripwave

dub45 wrote:

Probably become apparent - but when? Anyone know what decided Eircom to actually offer the product at this stage when there are so relatively few people still on bb?

75,000 DSL connections on 100 active exchanges? If 750 users per exchange is "relatively few", how many is "relatively a lot"?

Unfortunately, getting information out of eircom about what the situation is is like getting gold out of Fort Knox. But here probably are some exchanges where contention is happening. Some people mention their exchanges when reporting problems, but most don't. I think we need to start paying a lot more attention to exchanges the next time there's an outbreak of complaints about performance, and see if we can tell whether particular exchange are particularly problematic, and might be showing signs of exchange level contention.

dub45

Ripwave wrote:

75,000 DSL connections on 100 active exchanges? If 750 users per exchange is "relatively few", how many is "relatively a lot"?

380.000 based on this Article

Ripwave

dub45 wrote:

Well if 75,000 isnt relativley few why does everyone moan about the small number of bb users in Ireland?

75,000 BB connections in 1.2 million households is relatively few, but an average of 750 BB connections on a single exchange is more than enough to cause exchange level contention, if it's really an issue. (And it's in the nature of averages that some exchanges will have a lot more than 750, and some will have a lot less - though I suppose the exchanges that have a lot more are probably in areas where the exchange has significant amounts of data capacity, so it;s conceivable that exchange level contention isn't being enforced, just as the caps aren't).

dub45

Ripwave wrote:

75,000 BB connections in 1.2 million households is relatively few, but an average of 750 BB connections on a single exchange is more than enough to cause exchange level contention, if it's really an issue. (And it's in the nature of averages that some exchanges will have a lot more than 750, and some will have a lot less - though I suppose the exchanges that have a lot more are probably in areas where the exchange has significant amounts of data capacity, so it;s conceivable that exchange level contention isn't being enforced, just as the caps aren't).

So going back to my original question:) if as you say exchange level contention (possibly) isnt being eforced then the 24/1 is in fact a waste of money?

Just another point if the bandwidth keeps pace with the amount of subscribers then surely whether theres 75,000 or 3 million subcribers should not be an issue?

Ripwave

dub45 wrote:

So going back to my original question:) if as you say exchange level contention (possibly) isnt being eforced then the 24/1 is in fact a waste of money?

I didn't say it's not being enforced, I said that it's conceivable that it's not being enforced. It's conceivable that eircmo might start paying us to make phonecalls too, but I wouldn't be holding my breath.

But then, I don't know anyone who paid for a lower contention ration - I only kow of people who paid for a higher "download allowance" and got a lower contention ration as a side effect.

Just another point if the bandwidth keeps pace with the amount of subscribers then surely whether theres 75,000 or 3 million subcribers should not be an issue?

If you have 20 subscribers on an exchange, then they share a single 2Mb pipe, because that's the basic unit of connectivity used for exchanges. If you have 120 users on that exchange, they're still sharing that 2Mb pipe. When you get to 220 users, they'll be sharing 2x2Mb. When you're looking at those types of numbers, then, relatively speaking, there's always a lot of spare capacity. When you get to 2,000 users on an exchange, you're only up to 10x2MB - adding an extra 2MB when the next bunch of users come on isn't going to even be noticeable any more, and the relative amount of unallocated bandwidth is now tiny.

There's no point in using nation connection numbers when you're talking about exchange level contention. Most of the increase from 75,000 users to 350,000 users will come by increasing the number of connections on existing exchanges, not by adding new exchanges.

Ripwave

Ripwave wrote:

I didn't say it's not being enforced, I said that it's conceivable that it's not being enforced. It's conceivable that eircmo might start paying us to make phonecalls too, but I wouldn't be holding my breath.
.

<SNIP>

Its not a case of being enforced. It is to do with the ratio of bandwidth to people sharing it. If you happen to be on an exchange full of warez monkeys you may feel the contention in the form of lower up / down speeds but it is unlikely you will. For every warez monkey there are several light users.