What's wrong with Afrihost?

You've got a rather nasty flaw in your logic there & I wouldn't be surprised if it is the cause of recent troubles.

Network A

Moving 10 gigs per day
Capacity 20 gigs per day.
Utilizaiton: 50%
Chance of congestion: Low

Network B

Moving 14 gigs per day
Capacity 15 gigs per day.
Utilizaiton: 93%
Chance of congestion: High

Numbers pulled out of thin air of course...just demonstrating that gigs moved and capacity isn't linked.

i.e. You can't extrapolate capacity from gigs moved unless you also no the utilization %. Its like extrapolating a cars top speed from its current speed. Just doesn't work.

Give that man a Bells!
 
You've got a rather nasty flaw in your logic there & I wouldn't be surprised if it is the cause of recent troubles.

Network A

Moving 10 gigs per day
Capacity 20 gigs per day.
Utilizaiton: 50%
Chance of congestion: Low

Network B

Moving 14 gigs per day
Capacity 15 gigs per day.
Utilizaiton: 93%
Chance of congestion: High

Numbers pulled out of thin air of course...just demonstrating that gigs moved and capacity isn't linked.

i.e. You can't extrapolate capacity from gigs moved unless you also no the utilization %. Its like extrapolating a cars top speed from its current speed. Just doesn't work.

No offense, but thin air is right! :)

The amount of GB's per day show that we definitely have more capacity, which is measured per second on the IPC. Now bear in mind that IS ran out of IPC just before we left, so for the last month or so we were running at 100% capacity. So to move 23% more data, just doesn't need much more explanation, IMHO.

But at the end of the day, while you make a good point, why run the network at 50% if it means that some users are throttled so badly that they can't even open a web page. Why not run at 90% if it means that those downloading get a little extra (when capped users are catching some zzzz's) or however we can maximise the usage (without endangering performance) so that whatever benefit is realised we pass onto clients (at no extra cost)? And the "naughty" people get to browse and stream with decent speeds, while their torrents and nzb's are policed a little more strictly.
 
Now bear in mind that IS ran out of IPC just before we left, so for the last month or so we were running at 100% capacity.
Ah right. So IS IPC utilization is +- know after all. Then yes I'll agree that you can conclude on capacity based on gigs moved.

No offense, but thin air is right! :)
I included the single number available to me at that point: Increase in gigs %. The rest is by necessity thin air numbers. ;)

Broadly speaking, what I was getting at is that there is a disconnect between the info (More capacity, more gigs) and the actual user experiences (as witnessed by recent threads). It doesn't affect me personally right now, but I can't help but wonder wth is going on. There are only so many issues that it can be & the fact that both AH and Axxess are catching flak despite as you say different setups eliminates most of them. Only one common denominator left really. I'm sure it'll get sorted out eventually though.
 
Ah right. So IS IPC utilization is +- know after all. Then yes I'll agree that you can conclude on capacity based on gigs moved.


I included the single number available to me at that point: Increase in gigs %. The rest is by necessity thin air numbers. ;)

Broadly speaking, what I was getting at is that there is a disconnect between the info (More capacity, more gigs) and the actual user experiences (as witnessed by recent threads). It doesn't affect me personally right now, but I can't help but wonder wth is going on. There are only so many issues that it can be & the fact that both AH and Axxess are catching flak despite as you say different setups eliminates most of them. Only one common denominator left really. I'm sure it'll get sorted out eventually though.

I think it also requires some perspective - If we could get feedback from all of the tens of thousands of clients between us and axxess, we'd have a better picture of what is happening. But I think it's safe to say the MyBB users would be very different from Facebook users or Twitter users, while some may be all three. When we look at the numbers from our side, it makes a lot of sense. If you consider that top percentage of high bandwidth users consumers used like 30-40% more bandwidth, clearly that is being consumed by a small group of people (and they're probably not posting on any forums cos they know who they are).
 
But I think it's safe to say the MyBB users would be very different from Facebook users or Twitter users
Of course. We're [-]fussy[/-] sophisticated. :D

high bandwidth users consumers used like 30-40% more bandwidth, clearly that is being consumed by a small group of people (and they're probably not posting on any forums cos they know who they are).
I bet if you give rpm a nice detailed breakdown he could make an epic article out of it. *hint* *hint*
 
Top
Sign up to the MyBroadband newsletter
X