MWEB ADSL KPIs for Q1 2011 released

[rpm]

MWEB’s latest ADSL KPIs released

MWEB’s latest ADSL Key Performance Indicators show an improvement in both national and international performance

 

[snapple]

Am I reading the numbers correct? Packet loss is the only improvement; jitter is far worse (jitter is bad, right?) and so is national latency. Not sure where

significant improvement in network quality in both national and international performance

comes from.

 

[Lounger]

Average Round-trip latency (ms) Q3 2010=214 Q4 2010=235 Q1 2011=214
I don't see the improvement at all in international latency!

 

[ambo]

Average Round-trip latency (ms) Q3 2010=214 Q4 2010=235 Q1 2011=214
I don't see the improvement at all in international latency!

How much less are you realistically expecting? You'll shave less than 10ms off that if you were plugged directly into the router at the end of Seacom. SAT-3 is a little better but not by much. The speed of light is the limiting factor here.

 

[whatwhat]

How much less are you realistically expecting? You'll shave less than 10ms off that if you were plugged directly into the router at the end of Seacom. SAT-3 is a little better but not by much. The speed of light is the limiting factor here.

lol some people think that the speed of light increases every year by 10%, like inflation

 

[zeely]

dont forget that mweb stopped paying for peering somewhere in the middle, which forced a lot of companies to peer with them. this chart shows that things are pretty much back to normal, except mweb is no longer paying for peering

 

[Lounger]

How much less are you realistically expecting? You'll shave less than 10ms off that if you were plugged directly into the router at the end of Seacom. SAT-3 is a little better but not by much. The speed of light is the limiting factor here.

Do the math, 214-214 = 0 improvement.

 

[zeely]

Do the math, 214-214 = 0 improvement.

kpi's are ongoing measurement tool. who said we could expect an improvement? mweb never announced any improvements, what makes you think this quarters kpi's should differ from previous?

 

[guest2013-1]

Jitter is worse by a few hundred % and as it's pointed out, only latency really improved. Overall this is NOT a significant improvement.

 

[Bosvark@]

I would like to see that of IS and WA; and the rest of the self provisioning ISP's

 

[Conack]

I'm glad that Mweb's releasing their KPI's but what they're showing us here is pretty much only how much their service degraded over time and that it's now slowly getting back to what it was before, although not even entirely.

This is all fine and good news but lauding this recovery as "significant improvement in network quality" is a false claim according to their own numbers.

 

[Speedster]

kpi's are ongoing measurement tool. who said we could expect an improvement? mweb never announced any improvements, what makes you think this quarters kpi's should differ from previous?

Not sure if it was MWEB or the author who said

MWEB has now released their latest ADSL KPIs, showing a significant improvement in network quality in both national and international performance.

Either way, I think the issue isn't expecting improvements in latency etc - but the fact that an obviously incorrect (false) claim has been made.

 

[snobee]

Well, what I am happy about is that even though there isn't any real improvement, this is mweb we are talking about... the biggest uncapped provider in the country. Whether you love them or hate them, it shows their network at least has not gone down the tubes totally, and they are at least trying to sort out performance issues. So, if anything, possibly their network has improved considerably, as if it was the same as a year ago, with all those uncapped users, it should be completely up to crap (though some would say it is )

 

[Fudzy]

I would like to see that of IS and WA; and the rest of the self provisioning ISP's

Would be interesting to see but Mweb have been the only ones to provide their own stick to be beaten with.

 

[drukkie]

Well i think for the load increase mweb has experienced in the past year that this actually good news...

 

[Lounger]

Well i think for the load increase mweb has experienced in the past year that this actually good news...

I'm not sure I see your argument. A load increase normally is accompanied by a revenue and appropriate capacity increase, unless they are profit taking and decreasing quality (likely).

 

[OliverRC]

i.e. the reliability of packet times is less reliable and this is heralded as a "major improvement"

http://en.wikipedia.org/wiki/Packet_delay_variation

 

[groengras]

Average Round-trip latency (ms) Q3 2010=214 Q4 2010=235 Q1 2011=214
I don't see the improvement at all in international latency!

Blame Seacom,Telkom, any kind of com...

 

[Valerion]

I'm not sure I see your argument. A load increase normally is accompanied by a revenue and appropriate capacity increase, unless they are profit taking and decreasing quality (likely).

If their quality was decreasing, as you say, then it would have been accompanied by a worsening of the figures, or MWEB would not have published it (like the others). Latency has stayed the same, packet loss has decreased quite significantly. Jitter has increased. Overall I'd say the network is staying mostly stable, so not worsening.

And of course they take profit. Show me a for-profit telecomms company that does not.

 

[entrepr]

Jitter is worse by a few hundred % and as it's pointed out, only latency really improved. Overall this is NOT a significant improvement.

Jitter is mainly an issue if you're running a voip call or video call. Downloads such as youtube have a big enough buffer to cover the jitter numbers mentioned.

What I want to know is the 'uptime' kpi!