Variable speeds on Fibre Max

Forums › Spark New Zealand (including Skinny and BigPipe) › Variable speeds on Fibre Max

1 | 2 | 3

12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

#1787594 24-May-2017 17:11

alright, so that is seeming on par with a gbit connection - Obviously at max speeds that bin would be all of a few seconds if that.

What i'm wondering is if it is just the single testing server/speedtest client that is acting up..

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.

sidefx

3713 posts

Uber Geek

Trusted

#1787597 24-May-2017 17:20

kapitikarl:

It is noticeable with most online activities and given I am paying for fibre max and getting speeds only a fraction above their lower fibre package (namely fibre 200 - when I was getting 190+ consistently) I am keen to know what is causing this.

Absolutely agree you should get what you pay for, I just struggle to think of many online activities where the difference between 200Mbps and 800Mbps would be noticeable.

"I was born not knowing and have had only a little time to change that here and there." | Octopus Energy | Sharesies
- Richard Feynman

kapitikarl

151 posts

Master Geek

#1787599 24-May-2017 17:23

sidefx:

kapitikarl:

It is noticeable with most online activities and given I am paying for fibre max and getting speeds only a fraction above their lower fibre package (namely fibre 200 - when I was getting 190+ consistently) I am keen to know what is causing this.

Absolutely agree you should get what you pay for, I just struggle to think of many online activities where the difference between 200Mbps and 800Mbps would be noticeable.

Well Spark talked about a downgrade if they cant fix it. I suppose I am also thinking of those who access the network using wifi. Would there be any diff between 200 and 800 say for our tenant who lives on the edge of the wifi range?

hio77

12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

#1787600 24-May-2017 17:29

kapitikarl:

sidefx:

kapitikarl:

It is noticeable with most online activities and given I am paying for fibre max and getting speeds only a fraction above their lower fibre package (namely fibre 200 - when I was getting 190+ consistently) I am keen to know what is causing this.

Absolutely agree you should get what you pay for, I just struggle to think of many online activities where the difference between 200Mbps and 800Mbps would be noticeable.

Well Spark talked about a downgrade if they cant fix it. I suppose I am also thinking of those who access the network using wifi. Would there be any diff between 200 and 800 say for our tenant who lives on the edge of the wifi range?

Benefit would be Ethernet connected load not being as heavy on the other clients on connection.

Depending on how the wifi is actually setup, it could be noticeable (think 5ghz ac aps everywhere)

All depends on your expectations really.

Personally i'd pay for fibre max in a heartbeat - Two of them even! but that's just me and my network...

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.

sidefx

3713 posts

Uber Geek

Trusted

#1787674 24-May-2017 18:04

hio77:

Personally i'd pay for fibre max in a heartbeat - Two of them even! but that's just me and my network...

But then you are definitely not the average user :P

I had 200Mbps (before they started offering Max\Gbps) for a while but even when I decided to drop from that to 100 Mbps I realistically noticed almost zero difference. Only thing I really noticed was some Steam downloads took a bit longer. Not a big deal for me.

Your tenant is far more likely to have issues due to connecting via wifi than to notice a difference between 200 and 800 Mbps IMO...

"I was born not knowing and have had only a little time to change that here and there." | Octopus Energy | Sharesies
- Richard Feynman

l43a2

1779 posts

Uber Geek

ID Verified

Trusted

#1787684 24-May-2017 18:20

https://www.nperf.com/en/ another speedtest to try.

kapitikarl

151 posts

Master Geek

#1787689 24-May-2017 18:26

l43a2:

https://www.nperf.com/en/ another speedtest to try.

Ookla speed test - Ping 2ms; 259mbps dl (sigh); 290 mbps ul

nperf - 1.615ms; 861dl; 479ul

Now that is very odd. Ran ookla again and still report sub 300 dl.

Interestingly they are both using the same IP

Quic Broadband

Move to New Zealand's best fibre broadband service (affiliate link). Free setup code: R587125ERQ6VE. Note that to use Quic Broadband you must be comfortable with configuring your own router.

michaelmurfy

meow
13260 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#1787713 24-May-2017 19:12

kapitikarl:

Interestingly they are both using the same IP

The IP you're seeing is your own, not the Speedtest server. Spark do not have any testing servers on npref's platform.

Looks like to me that you're just seeing congestion on the testing server itself - your connection won't be affected and any feeling of slowness is likely due to a pure placebo effect. It also may be something else is downloading on your network so the only real way of testing this theory is to boot into another OS (like Linux) on your machine, disconnect everything from your router including wireless devices and do a few tests under a live distro to confirm there is also nothing on your PC slowing it down.

Michael Murphy | https://murfy.nz
Referral Links: Quic Broadband (use R122101E7CV7Q for free setup)

Are you happy with what you get from Geekzone? Please consider supporting us by subscribing.
Opinions are my own and not the views of my employer.

sbiddle

30853 posts

Uber Geek

Retired Mod

Trusted

Biddle Corp

Lifetime subscriber

#1787807 24-May-2017 21:02

sidefx:

I just struggle to think of many online activities where the difference between 200Mbps and 800Mbps would be noticeable.

That's because there aren't any.

yitz

2081 posts

Uber Geek

#1787811 24-May-2017 21:06

Perhaps gigabit would be good for running a speed test server on the IP of your broadband connection

kapitikarl

151 posts

Master Geek

#1787815 24-May-2017 21:14

So let me get this right - a reduction in speed is all in my mind and I don't need the speed anyway - have I got that right :p

Talkiet

4793 posts

Uber Geek

Trusted

#1787823 24-May-2017 22:13

Speedtests are unreliable at speeds approaching gig, especially when you take into account all the other factors and things that have to be _EXACTLY_ right for consistent speeds over 900Mbps to be reported. You are almost certainly measuring variability in the speedtest results, not a variance in the speed of your service.

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

kapitikarl

151 posts

Master Geek

#1787824 24-May-2017 22:19

All good mate. I have closed the case. It seems ookla is woefully inaccurate compared to the other speed tests.I have uninstalled it and obtaining a frontal lobotomy for my delusional thinking lol.

hio77

12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

#1787825 24-May-2017 22:20

Talkiet:

Speedtests are unreliable at speeds approaching gig, especially when you take into account all the other factors and things that have to be _EXACTLY_ right for consistent speeds over 900Mbps to be reported. You are almost certainly measuring variability in the speedtest results, not a variance in the speed of your service.

Cheers - N

This reminds me to run more tests on my 10G machines overseas... 20G peering to speedtest server in the same datacenter and the speedtest client is just dog slow still..

now raw wgets... oh boy..

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.

Talkiet

4793 posts

Uber Geek

Trusted

#1787826 24-May-2017 22:25

Yeah, our speedtest servers aren't slow... This is a wget from a $50 odroid on my home network from one of the Spark speedtest servers... (obviously on Spark Fibre Max)

OK, that image isn't working for me for a reason I don't care to try and troubleshoot... try this - http://neilnz.com/stuff/fastodroid.jpg

The ookla app is designed to give a reasonable idea of achievable peak performance but it does it without saturating the link for long, and it does it adaptively by adjusting the actual test depending on a number of factors it tries to determine at the beginning of the test. Although it's often fine, it frequently gives misleading results depending on browser, OS, network setup etc...

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

1 | 2 | 3