TrueNet testing suggests Southern Cross Cable is roadblock on UFB performance

Forums › New Zealand Broadband › TrueNet testing suggests Southern Cross Cable is roadblock on UFB performance

lchiu7

6475 posts

Uber Geek

Trusted

#119535 5-Jun-2013 09:00

See in NBR and originated by Truenet.

http://www.nbr.co.nz/article/truenet-testing-suggests-southern-cross-cable-roadblock-ufb-performance-ck-141043

Results are somewhat surprising but also a ring of truth in them.

Certainly I never found TCL 100Mbs cable much faster overseas than my more modest 15Mbs TCL or 13Mbs Telecom xDSL except when I can create multiple streams of connections with say BT.

Then TCL 100Mbs is pretty fast.

But for video streaming overseas, it was hardly worth the cost.

Staying in Wellington. Check out my AirBnB in the Wellington CBD. https://www.airbnb.co.nz/h/wellycbd PM me and mention GZ to get a 15% discount and no AirBnB charges.

1 | 2

1469 posts

Uber Geek

ID Verified

#830814 5-Jun-2013 09:03

Of course 100mbit HFC is faster than xDSL :P Whats interesting is it's slightly better than UFB 100mbit which makes me wonder first off if Vodafone's caching is helping here or if the UFB 100mbit circuit ISP wasn't upto scratch

Most problems are the result of previous solutions...

All comment's I make are my own personal opinion and do not in any way, shape or form reflect the views of current or former employers unless specifically stated

Sounddude

I fix stuff!
1928 posts

Uber Geek

Trusted

2degrees

Lifetime subscriber

#830815 5-Jun-2013 09:07

Latency is always going to have an effect on download times.

its not fair to say that SCCN is the issue. The issue is we are so far away from the content.

chevrolux

4962 posts

Uber Geek
Inactive user

#830857 5-Jun-2013 10:51

I don't get why RBI has been lumped with ADSL. Any ADSL cabinets that are being upgraded in the RBI project get ADSL2+/VDSL line cards so should be considered in the same group as the other ADSL probes.
And what is going on with Vodafone's Australia download times? Have they got a screwed up route or something?

NonprayingMantis

6434 posts

Uber Geek

#830859 5-Jun-2013 10:54

chevrolux: I don't get why RBI has been lumped with ADSL. Any ADSL cabinets that are being upgraded in the RBI project get ADSL2+/VDSL line cards so should be considered in the same group as the other ADSL probes.
And what is going on with Vodafone's Australia download times? Have they got a screwed up route or something?

the RBi might be referring to the VF fixed-wireless solution,

JamesL

956 posts

Ultimate Geek
Inactive user

#830860 5-Jun-2013 10:56

Sounddude: Latency is always going to have an effect on download times.

its not fair to say that SCCN is the issue. The issue is we are so far away from the content.

This

mercutio

1392 posts

Uber Geek

#830862 5-Jun-2013 10:58

the graphs clearly show that telstraclear australia performance is shocking.

NZCrusader

646 posts

Ultimate Geek

#830939 5-Jun-2013 13:21

Sounddude: Latency is always going to have an effect on download times.

its not fair to say that SCCN is the issue. The issue is we are so far away from the content.

Depends on what type of traffic it is.

Will certainly affect TCP, but not UDP so much.
A delayed UDP stream is just that. Delayed. It doesn't actually have an effect on the throughput.
TCP delays will affect throughput however.

There is actually nothing wrong with southern cross other than the ISPs are not purchasing enough bandwidth. Which in turn is affected by the price per megabit.

Really speaking, until we can get some more international links (similar to the excessive quantity across the Atlantic) then there is not enough competition to drive down prices and compel ISPs to switch from heavy caching investments to just purchasing more bandwidth to meet customer demand.

Besides another factor is the capable speed of the server on the other end, their network policies and providers.

NZ / AU Battlefield 4 Gaming Community
http://www.sonsofvalour.net/forums/forum.php

GoodSync

GoodSync. Easily back up and sync your files with GoodSync. Simple and secure file backup and synchronisation software will ensure that your files are never lost (affiliate link).

Ragnor

8219 posts

Uber Geek

Trusted

#831242 5-Jun-2013 20:16

The headline was clearly sensationalized for effect when compared to what John was quoted as saying in the article.

insane

3239 posts

Uber Geek

ID Verified

Trusted

#831252 5-Jun-2013 20:30

Wow Truenet are not doing their reputation any good with these pathetic tests, but mainstream media will undoubtedly love it. That's twice in a month now, remember those VoIP tests...

Ask any provider who purchases bandwidth from SCCN, you get exactly what you pay for, they actually offer a damn good service.

And testing high-speed connections using 300-500KB websites... OK, must be accurate, read it on the internet.

mercutio

1392 posts

Uber Geek

#831256 5-Jun-2013 20:33

insane: Wow Truenet are not doing their reputation any good with these pathetic tests, but mainstream media will undoubtedly love it. That's twice in a month now, remember those VoIP tests...

Ask any provider who purchases bandwidth from SCCN, you get exactly what you pay for, they actually offer a damn good service.

And testing high-speed connections using 300-500KB websites... OK, must be accurate, read it on the internet.

it is real world in a way. it's best to have low latency low packet loss, fast dns, local caching in order to get best performance. of course truenet do cache busting unlike proper regular sites. (a few dodgy sites purposefully force cache disabling on some of their stuff)

mercutio

1392 posts

Uber Geek

#831263 5-Jun-2013 20:36

NZCrusader:
Sounddude: Latency is always going to have an effect on download times.

its not fair to say that SCCN is the issue. The issue is we are so far away from the content.

Depends on what type of traffic it is.

Will certainly affect TCP, but not UDP so much.
A delayed UDP stream is just that. Delayed. It doesn't actually have an effect on the throughput.
TCP delays will affect throughput however.

i dunno why everyone still seems to think it's delay that effects bandwidth on high latency connections. it's packet loss. if you have 0% packet loss you can get line rate to overseas. if you have even minor packet loss it can severely degrade performance.

for the same level of packet loss, the less latency the better it'll handle it.

but with modern tcp improvements like cubic congestion control it's not hard to do 100 megabit+ to the other side of the world.

dwl

371 posts

Ultimate Geek

#831485 6-Jun-2013 10:25

I agree that packet loss controls the available rate for TCP but latency still has a major effect. High latency needs lower loss for the same throughput. For all major ISPs I suggest there is probably some low level of loss on SCCN at busy times and I can't believe they would ever buy enough capacity to let all concurrent users get their full access rates.

There is a relationship with local access type on comparative performance. For UFB there may only be a few milliseconds delay to local servers so adding at least 25 ms (or much more to the US) is proportionally a much bigger difference. For ADSL if we start with say 35ms local access delay then going across SCCN to Sydney isn't so much of a change. The total latency from Auckland to Sydney on ADSL may easily be twice that on UFB (which for same underlying SCCN low level loss translates to different TCP rates if the ISP isn't limiting).

The use of web page load times is an interesting metric but I think we risk adding too much complexity and risk making incorrect assessments about SCCN depending on what processing is done by ISPs. The Truenet test page seems to have 33 page elements that need to be downloaded. Using Firefox and Firebug on Windows 8 on Telecom ADSL I see blocking that seems to limit to 6 concurrent threads and much higher total page load times than I can get using Linux wget for all elements.

I really support what Truenet is trying to achieve, identifying performance issues across multiple providers and technologies and trying to relate to the real world with web page load times. Sounddude is right that we are a long way from the content but I think SCCN is a limitation driven by valid commercial pressures. I'm just not sure web page load times is the best metric but this depends on how well an automated test aligns with user experience. Any other thoughts out there?

ubergeeknz

3344 posts

Uber Geek

Trusted

Vocus

#831490 6-Jun-2013 10:35

mercutio:
NZCrusader:
Sounddude: Latency is always going to have an effect on download times.

its not fair to say that SCCN is the issue. The issue is we are so far away from the content.

Depends on what type of traffic it is.

Will certainly affect TCP, but not UDP so much.
A delayed UDP stream is just that. Delayed. It doesn't actually have an effect on the throughput.
TCP delays will affect throughput however.

i dunno why everyone still seems to think it's delay that effects bandwidth on high latency connections. it's packet loss. if you have 0% packet loss you can get line rate to overseas. if you have even minor packet loss it can severely degrade performance.

for the same level of packet loss, the less latency the better it'll handle it.

but with modern tcp improvements like cubic congestion control it's not hard to do 100 megabit+ to the other side of the world.

Latency definitely affects TCP bandwidth, in most cases. Yes some very recent TCP stacks might address this somewhat, but most equipment is not running the latest TCP stacks.

dwl

371 posts

Ultimate Geek

#831504 6-Jun-2013 11:04

We need some clarity about what typical TCP stacks are being used and what "recent TCP stacks" means. Improved error correction has been around for many years and I believe appeared with Windows XP which is now starting to drop off as the most common OS. I don't believe the very latest stacks are essential, simply one that has window scaling and SACK (which I think should be many).

I think many users could get high bandwidth if the path to the originating server had low enough loss. It isn't fair to blame only SCCN bottlenecks as many other segments, outside NZ ISP control, will also play a key role but I don't think it is valid to suggest the limitation is necessarily the user or server TCP stacks.

NonprayingMantis

6434 posts

Uber Geek

#831514 6-Jun-2013 11:13

insane: Wow Truenet are not doing their reputation any good with these pathetic tests, but mainstream media will undoubtedly love it. That's twice in a month now, remember those VoIP tests...

Ask any provider who purchases bandwidth from SCCN, you get exactly what you pay for, they actually offer a damn good service.

And testing high-speed connections using 300-500KB websites... OK, must be accurate, read it on the internet.

to be fair to truenet, they dont state anything about sccn being the problem in their report, that is Chris Keal's interpretation. Sounds liek he probed John butt about whether southern cross effects international website access and the answer 'well, I guess it partly effects it" and Keal took that to mean sccn is an articificial profit driven bottleneck when really it isn't, it's a real world physics bottleneck.

1 | 2