[TelstraClear] 25mbps roll out?

Forums › One NZ (including Farmside) › [TelstraClear] 25mbps roll out?

1 | 2 | 3 | 4 | 5 | 6

amanzi

Amanzi
1355 posts

Uber Geek
+1 received by user: 335

ID Verified

Trusted

Lifetime subscriber

#125910 23-Apr-2008 15:46

jpollock: Actually, I don't understand why they don't give out access to the speedtest.

I was just thinking the same thing...

jamgol

42 posts

Geek

#125912 23-Apr-2008 15:53

They need a speed test site that is reliable and not overloaded ..... especially when they are doing it in front of business and residential customers. It needs to be totally consistant and therefore dedicated to them.

Same software as used here: (not sure where this is hosted though)
http://www.consumerspeedtest.org.nz/speedtest.php

The techs were saying that they get heaps of people logging faults because "speedtest.net" said their connection is slower than it was yesterday. I can imagine it would be 10 times worse if it was provided by them, "my connection is 512K slower than ten minutes ago, can you fix it ?"

Unfortunately for us, most Telstra customers don't understand the numbers, so it would just make their already overloaded call centers even worse.

I must admit, It would be nice though. Any chance of a password Telstra ? .... just for geekzone users only ?

jamgol

42 posts

Geek

#126120 24-Apr-2008 13:34

Don't know what they've done but speeds are great today .... I think I love Warpspeed already.

Download

doppleganger

33 posts

Geek

#126275 25-Apr-2008 10:29

jamgol:Same software as used here: (not sure where this is hosted though)
http://www.consumerspeedtest.org.nz/speedtest.php

That would be on the iconz network, which appears to be directly off of TCL with some pretty good latency:

mrlappy:~ $ host www.consumerspeedtest.org.nz
www.consumerspeedtest.org.nz is an alias for consumerspeedtest.org.nz.
consumerspeedtest.org.nz has address 210.48.76.148

mrlappy:~ $ traceroute 210.48.76.148
traceroute to 210.48.76.148 (210.48.76.148), 64 hops max, 40 byte packets
1 * * * * * 1.081 ms 0.704 ms 0.535 ms
2 * * * * * 14.960 ms 18.027 ms 15.339 ms
3 * * * * * 26.559 ms 20.715 ms 18.873 ms
4 pe14-bertha.paradise.net.nz (203.96.155.226) 13.936 ms 14.952 ms 20.767 ms
5 g1-0-1042.u21.tar.telstraclear.net (203.98.23.41) 24.624 ms 17.178 ms 19.059 ms
6 a12-2-23.u21.tar.telstraclear.net (203.98.23.33) 20.111 ms 21.520 ms 17.652 ms
7 gi-4-0-878.u11.telstraclear.net (218.101.61.137) 27.679 ms 30.474 ms 31.047 ms
8 gi-9-0-897.u12.engops.telstraclear.net (203.98.23.22) 31.013 ms 35.643 ms 32.875 ms
9 lnk-iconz-g1-0-1396.u12.brfd.telstraclear.net (203.98.18.14) 30.560 ms 29.182 ms 26.322 ms
10 ge-0-2-0.sky01-jcore.iconz.net (210.48.5.69) 29.456 ms 40.642 ms 33.772 ms
11 ge-0-3-0.air01-jcore.iconz.net (210.48.5.2) 35.698 ms 30.988 ms 29.205 ms
12 ge-1-1.cat01-air01.iconz.net (210.48.5.102) 28.339 ms 27.527 ms 27.174 ms
13 ip-210-48-76-148.iconz.net.nz (210.48.76.148) 27.526 ms 25.128 ms 25.754 ms

Opinions expressed by me are mine alone and should not be construed to be supported by my employer in any way shape or form. Ever. Even if they say they agree. Especially if they say they agree. Really!

sbiddle

30853 posts

Uber Geek
+1 received by user: 9996

Retired Mod

Trusted

Biddle Corp

Lifetime subscriber

#126278 25-Apr-2008 10:47

/me personally thinks that speedtests should all be pulled from the interw3b. As pointed out by numerous people their results can be absolutely meanless, particularly the results for many of the NZ based ones.

doppleganger

33 posts

Geek

#126279 25-Apr-2008 11:05

sbiddle: /me personally thinks that speedtests should all be pulled from the interw3b. As pointed out by numerous people their results can be absolutely meanless, particularly the results for many of the NZ based ones.

I don't think they need to be pulled at all. What they need is info to let the people using them evaluate the results realistically.

They would need:

1) Max available bandwidth,
2) Current utilisation, preferably graphed against time,
3) Accurate latency to host, plus current latency to static net hosts again graphed against time
4) Ability to select multiple tcp steams from 1 to say 5... Who has more than 5 regular simultansous single tcp stream downloads out there?
5) Aggregate bandwith reporting across multiple streams,
6) Much info about how TCP and the internet in general actually works.

Anyone got any other ideas on how to make bandwith reporting realistic?

HP

Shop now for HP laptops and other devices (affiliate link).

BarTender

3629 posts

Uber Geek
+1 received by user: 2572

ID Verified

Trusted

Lifetime subscriber

#126304 25-Apr-2008 14:03

doppleganger: Anyone got any other ideas on how to make bandwith reporting realistic?

MTRG realtime reporting against traffic flows between your router and Cable modem. Graph it over at least 2-3 week period, collect information about TCP response times as well as throughput. Then compare against other ISP's or other customers on the same plan doing similar traffic.
Speed tests are pretty irrelevant IMHO as well, having the TCL Speedtest site would be ideal, since then you know you are correctly configured with the right up&down speeds, otherwise I see no point in wasting your data cap on a speed test, you know when it's running slowly :)

1 | 2 | 3 | 4 | 5 | 6