[AusNOG] NBN/ACMA fines etc

Ross Wheeler ausnog at rossw.net
Tue Jan 2 15:31:19 EST 2018

On Tue, 2 Jan 2018, paul+ausnog at oxygennetworks.com.au wrote:

> Yes there should be transparency but it should not be something that an end user should be looking at

Not sure I completely agree with you there.

> NBN have broken so much

Could have left it at that :)

> this is why we now have such a big mess with speeds and expectations.

I believe most of it comes back to the thing being marketed and sold as
"superfast", with promises of "100 megabits" ("up to" came later, and was 
typical legal-speak in tiny print that few people got, and none of us 
SHOULD have needed to see)

> Unless you overprovision the layer 2 connection an end user will never 
> see full speed.

Even if you DO overprovision, the end user is never guaranteed of full 
speed, due to there being significant chunks that are out of the ISP/RSP 
reach and visibility.

> There are so many variables in these things now that it's next to 
> impossible to point the finger, especially if you are an end user,

Yes, and most of the players seem(ed) to love it. Everyone pointed the 
finger at someone else and nobody had to "fix" it.

If I might add my own personal experience by way of example.
Where I live I cannot get FTTN. I cannot get FTTP. My choices are 
basically satellite (yeah, right) or fixed wireless. Fixed wireless was 
duly provisioned after a lot of kicking and screaming (another story in 
itself) and my service was delivering speeds that averaged about 10Mbps 
down and 4Mbps up over a 24-hour period, but with speeds around 0.7Mbps 
down and 4Mbps up from most of the time between 3pm and midnight.

The RSP had over-provisioned CVC to the POI I was (indirectly) connected 
to. Faults were raised, things were checked. The service was deemed to be 
"operating within acceptable limits". A second service was provisioned 
through an entirely different RSP, different carrier, different CVC etc. 
Similar (slightly worse) performance was observed. The significant speed 
drops (and subsequent recoveries) were exactly co-incident across both 

A third RSP was run. Not surprisingly, again, virtually identical results.
Complaints made, letters with tests and graphs were sent to local MPs, the 
ACMA, ACCC and nbnco. NBN came and changed me to a different cell on the 
tower (changed equipment from 2.5 to 3.4GHz). The results were about a 10% 
improvement over a 24-hr period, but the 8+ hour periods of unusability 
with speeds down to 0.5Mbps continued, and over the next 6.5 months got 
progressively worse, unti in early December the 24-hour average throughput 
dropped to under 5Mbps. More strongly worded complaints... and perhaps no 
coincidently, the ACCC getting a bit noisy about things... and my 50/20 
service suddenly changed 2 weeks ago to one that is delivering 25-30Mbps 
average down and 6-8Mbps up) with "peak times" usage now holding at around 

This is not to say that it is by any measure "acceptable", but it 
certainly shows that the 5 end-to-end RF links that nbn are using to 
connect this regional tower to the chosen POI over 100km away is, and has 
always been, seriously over-contended. Worse, it is in a part of the 
network that the RSPs claim no visibility of and no control over.

If nbn are incapable of (or unwilling to) admit when there are problems 
and address them in a timely manner, then there should be a legislated, 
enforcable (and enforced) way of making that information available, and a 
mechanism to ensure nbn deliver a service that is actually "fit for 


More information about the AusNOG mailing list