|
|
|
I'd get a 300/100 connection with an RSP with a local POP and good peering, and understand that routes change often, usually outside of their network so not always within their control.
I'm still not sure what you're trying to show here? If you intend on running your connection full tilt while gaming you will obviously need to upgrade your network equipment as your LAN only appears to be gigabit Ethernet... 900/500 connections are shaped nicely to run within the envelope of GE while hyperfibre will be more jagged as your GE is maxed out so it's likely your problem with high upload latency under load is occurring before things even hit the Chorus or ISP network.
From my observation in recent weeks Compass appear to have completed some network augments as I'm seeing much better international performance even on regular fibre. Most of their external traffic goes via Voyager who have decent responses times to most international destinations. They have national peering and CDNs over Megaport/AKL-IX across Auckland locations including 220 Queen, Mayoral Drive and their own Datavault. Since you're also in Auckland any further local peering (like in WLG/CHC) won't make a difference.
What would you all do these days if your priority is low latency to game servers? Consider that 2dgrees uses CGNAT, and they and MyRepublic have now both been bought into the same group.
There is simply no such thing. MyRepublic was marketing and not a good ISP and everyone else are pretty much of a muchness.
What are you expecting? Is there actually any problem here...? You can still game fine...
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.
Does CGNAT really affect latency? Most ISP's use it these days with an option to opt out.
Wombat1:I would say it can... it's an extra hop, recall the days of transparent proxy caches where the load balancers on a bad day would add 100's of milliseconds or stall on connection setup. It can also depend on network topology - I recall at some stage either on Snap or Devoli that CG-NAT boxes and the BNGs were in different cities. So say you are in Christchurch and the CGNAT box for Christchurch customers was in Wellington so CGNAT connections across town had to take the round trip to Wellington whereas publicly routed would be directly across town.
Does CGNAT really affect latency? Most ISP's use it these days with an option to opt out.
I don't think Compass have CG-NAT though, could be wrong.
This thread is more about Hyperfibre ISPs either providing or the customers themselves supplying their own router with 2.5 Gbps WAN port but 1 Gbps LAN ports then the customer experience will be suboptimal if they saturate their 1 Gb LAN.
havenplaygame:
What would you all do these days if your priority is low latency to game servers?
Stick with regular fibre and be grateful I no longer live in the days of dial-up internet where an average ping to a local game server was over 200ms?
michaelmurfy:
Yeah more marketing cruft. The other problem is you've change from a large, well peered ISP to a smaller provider. But to be honest you're highly unlikely to notice any difference even with online gaming... You're changing for all the wrong reasons if you only care about latency anyway.
It's even worse if you scroll down to FAQs about Hyperfibre:

Good god, who's making that claim?
Cheers - N
[edit: I see. Zeronet. I guess "Zero" refers to the amount of time in seconds given over to critical analysis of that claim]
backfiah:
michaelmurfy:
Yeah more marketing cruft. The other problem is you've change from a large, well peered ISP to a smaller provider. But to be honest you're highly unlikely to notice any difference even with online gaming... You're changing for all the wrong reasons if you only care about latency anyway.
It's even worse if you scroll down to FAQs about Hyperfibre:
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.
@BMarquis - are you able to provide the Chorus testing results showing a 6-50x improvement in latency that can be achieved moving from "standard Fibre internet" to Hyperfibre please?
(I kid, I kid - but perhaps a comment distancing yourselves from this sort of creative marketing might be nice :-)

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.
Chorus found a way to speed up the speed of light? Amazing....
They must've been comparing with FibreX 😉
Talkiet:
@BMarquis - are you able to provide the Chorus testing results showing a 6-50x improvement in latency that can be achieved moving from "standard Fibre internet" to Hyperfibre please?
(I kid, I kid - but perhaps a comment distancing yourselves from this sort of creative marketing might be nice :-)
Given xgspon is being used for HPA, it might not be as silly of a question if you consider it's TDMA in the uplink.
hyperfibre is the wrong context to ask that question, to be fair...
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
To be honest if there was a "layer 3" variant of Hyperfibre where Chorus handled the CG-NAT and you had two users in Invercargill directly exchanging traffic with each other you could easily make the 50ms claim that compared to government mandated layer 2 "standard fibre" which potentially trombones traffic up to Auckland two rounds trip for tail extensions and national peering.
Lorenceo:
NBN's artificial 50Mbps upload limit is just not up to snuff in 2023.
Wombat1:
This is not correct. All FTTP addresses in Aus are capable of speeds up to 1000/400 Mbps, they just expensive and not the norm like in NZ
|
|
|