Geekzone: technology news, blogs, forums
Guest
Welcome Guest.
You haven't logged in yet. If you don't have an account you can register now.


View this topic in a long page with up to 500 replies per page Create new topic
1 | 2 | 3

edc

edc

31 posts

Geek


  #2235851 13-May-2019 11:30
Send private message

HowickDota: Have you tried playing without the ping showing? I've played Dota for 11+ years and always adapted to the ping, from LAN up to 240ms. I wonder if you did a blind test between 33 and 43ms would you be able to guess correctly 9/10 times

 

I noticed the issue while playing, not by looking at latency. 




noroad
1037 posts

Uber Geek
+1 received by user: 691

Trusted

  #2235885 13-May-2019 12:15
Send private message

It depends which NZ-AU cable your ISP is using, Minimum times are around - Southern Cross=23ms, TGA=25ms, Hawaiki=35ms to Auckland. Then add transit time from you to Auckland.

 

 

 

 

 

 

 

 


BarTender
3630 posts

Uber Geek
+1 received by user: 2572

ID Verified
Trusted
Lifetime subscriber

  #2235891 13-May-2019 12:22
Send private message

Lias:

 

Do any of you guys saying 10ms additional latency is irrelevant actually play twitch based shooters at a competitive level? Because it absolutely can matter.

 

 

Being a late player to this thread I think what is being consistently said. If you want a decent ping and routing then stick with a decent ISP. 

 

And that outside of your internal network (using properly terminated wired connection, decent router capable of shifting the traffic as did anyone else notice pfsense in the dns name of the first hop?) there isn't much you can do.

 

If it were me, I would swap to an ISP supplied router rather than using pfsense just to test that, plus make sure there wasn't any other traffic possibly saturating the connection or router.

 

Then if it was still bad I would ask others in my local region who are with other ISPs and see. As the ping in Auckland vs Wellington vs Christchurch to Sydney or further away is +10 (WLG) or +25 (CHC) ms or so.




edc

edc

31 posts

Geek


  #2235895 13-May-2019 12:32
Send private message

noroad:

 

It depends which NZ-AU cable your ISP is using, Minimum times are around - Southern Cross=23ms, TGA=25ms, Hawaiki=35ms to Auckland. Then add transit time from you to Auckland.

 

 

That's useful, do you happen to have a table that details it per ISP, or know which MyRepublic use? 


darylblake
1175 posts

Uber Geek
+1 received by user: 413

Trusted

  #2235897 13-May-2019 12:34
Send private message

Minimum ping Auckland -> SYD is about 24ms.

 

In gaming you will not notice anything between 50ms and 25ms.... So don't fret it. 

 

Even if you were a pro gamer, playing an FPS game you would need about 100ms+ to notice a lag. 

 

Its more than likely any lag you encounter will be to either your PC catching up on processing (running background tasks which are causing a lag). Or frames being back-generated / catching up. Or you are using wireless (which is a terrible idea for games).

 

Don't complain about a 10 ms difference on fibre. Thats just ridiculous.


hio77
'That VDSL Cat'
13044 posts

Uber Geek
+1 received by user: 3901

ID Verified
Trusted
Lizard Networks
Subscriber

  #2235902 13-May-2019 12:40
Send private message

edc:

 

noroad:

 

It depends which NZ-AU cable your ISP is using, Minimum times are around - Southern Cross=23ms, TGA=25ms, Hawaiki=35ms to Auckland. Then add transit time from you to Auckland.

 

 

That's useful, do you happen to have a table that details it per ISP, or know which MyRepublic use? 

 

 

i'd suspect they have diversified and purchased another route, your traffic happens to be across that.

 

 

 

It's extremely rare that any provider would favor latency over all the other metrics that can account for a better route.

 

truth be told, 99% of people prefer consistency and faster speed over 10ms of latency. 

 

 

 

Also remember while on paper X Y cable has Z latency, you need to account for what goes on within your provider, or the backhaul through to your connection.

 

 

 

Again, lets take chorus for example. Their main metric is ensuring a link isn't congested, not that it's the direct path (although it typically makes up maybe 1ms difference)

 

The only way you can ensure you get that 24ms to sydney is to have a dedicated, agreement with the provider providing it to provide you X route.





#include <std_disclaimer>

 

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


 
 
 

Shop now on AliExpress (affiliate link).

edc

edc

31 posts

Geek


  #2235906 13-May-2019 12:42
Send private message

BarTender:

 

Lias:

 

Do any of you guys saying 10ms additional latency is irrelevant actually play twitch based shooters at a competitive level? Because it absolutely can matter.

 

 

Being a late player to this thread I think what is being consistently said. If you want a decent ping and routing then stick with a decent ISP. 

 

And that outside of your internal network (using properly terminated wired connection, decent router capable of shifting the traffic as did anyone else notice pfsense in the dns name of the first hop?) there isn't much you can do.

 

If it were me, I would swap to an ISP supplied router rather than using pfsense just to test that, plus make sure there wasn't any other traffic possibly saturating the connection or router.

 

Then if it was still bad I would ask others in my local region who are with other ISPs and see. As the ping in Auckland vs Wellington vs Christchurch to Sydney or further away is +10 (WLG) or +25 (CHC) ms or so.

 

 

I've tried two pfsense boxes, pings tested when the link was idle, Sunday night around 11 pm. The pings are constant 42ms no matter the time of day though, I've tested in evenings every day of the week. I have two R7000s to test on the link, but I really don't think it can make a difference because the delay is between hops outside of my network. I use CAT6a 550 MHz cables, I think the cable is 3m in length. The router I have now has an Intel i350 T4V2 (https://ark.intel.com/content/www/us/en/ark/products/84805/intel-ethernet-server-adapter-i350-t4v2.html). I can see with the traceroutes that Auckland to Sydney should be about 23 ms. The pfsense setup is essentially out of the box, I only activated the 2 additional NICs. 


noroad
1037 posts

Uber Geek
+1 received by user: 691

Trusted

  #2235912 13-May-2019 12:46
Send private message

edc:

 

That's useful, do you happen to have a table that details it per ISP, or know which MyRepublic use? 

 

 

 

 

I couldn't answer that on behalf of others sorry as its really up to the individual ISP's to decide if they want to release that info. Here is real world ping times between core transit routers (as good as it possibly gets) for reference.

 

 

 

Southern Cross -

 

64 bytes from 10.93.10.39: icmp_seq=1 ttl=64 time=23.1ms.
64 bytes from 10.93.10.39: icmp_seq=2 ttl=64 time=23.0ms.
64 bytes from 10.93.10.39: icmp_seq=3 ttl=64 time=23.0ms.
64 bytes from 10.93.10.39: icmp_seq=4 ttl=64 time=23.0ms.
64 bytes from 10.93.10.39: icmp_seq=5 ttl=64 time=23.3ms.

 

 

 

TGA -

 

64 bytes from 10.93.10.37: icmp_seq=1 ttl=64 time=25.9ms.
64 bytes from 10.93.10.37: icmp_seq=2 ttl=64 time=25.9ms.
64 bytes from 10.93.10.37: icmp_seq=3 ttl=64 time=25.9ms.
64 bytes from 10.93.10.37: icmp_seq=4 ttl=64 time=25.9ms.
64 bytes from 10.93.10.37: icmp_seq=5 ttl=64 time=25.9ms.

 

 

 

I don't have Hawaiki so can't give you that one.

 

 

 

 

 

 

 

 


edc

edc

31 posts

Geek


  #2235913 13-May-2019 12:48
Send private message

darylblake:

 

Minimum ping Auckland -> SYD is about 24ms.

 

In gaming you will not notice anything between 50ms and 25ms.... So don't fret it. 

 

Even if you were a pro gamer, playing an FPS game you would need about 100ms+ to notice a lag. 

 

Its more than likely any lag you encounter will be to either your PC catching up on processing (running background tasks which are causing a lag). Or frames being back-generated / catching up. Or you are using wireless (which is a terrible idea for games).

 

Don't complain about a 10 ms difference on fibre. Thats just ridiculous.

 

 

As I've stated many times, I do notice the difference post change. The ping is measured consistently from the router, in client (in game). It isn't wireless (I'm cabled), it isn't Windows, it isn't my hardware. Nothing has changed apart from the ISP's performance.

 

There are many aspects of lag, the link being one of them, and the additional 10ms over and above what already exists, is bothersome. 


TheoM
229 posts

Master Geek
+1 received by user: 55

ID Verified
Trusted

  #2235914 13-May-2019 12:49
Send private message

edc:

 

[...] I can see with the traceroutes that Auckland to Sydney should be about 23 ms. The pfsense setup is essentially out of the box, I only activated the 2 additional NICs. 

 

 

Go get yourself a cheap HG659/b and test using that. pf really isn't very useful for diagnosing anything. Anything from BSD's drivers for your NIC to the fact BSD's implementation of PPPoE is single threaded can make a huge difference.

 

While Auckland to Sydney can be 24ms, that is getting very close to the theoretical minimum RTT ICMP Echo time. A lot can influence that ping, from your internal routing configuration to the routers on the way (for example, since ICMP Echo is the lowest priority, if a router is busy enough it can elect to ignore the ping, or delay it until it has enough time to respond). 

 

As was said previously, it also depends on exactly which cable your traffic flows over.





Hi! I'm TheoM, but you know that already. I run Linux mirrors in NZ together with 2degrees. Like a mirror added? PM me!

 


 

https://theom.co.nz | https://theom.nz | https://mirrorlist.mirrors.theom.nz | Providing Free Mirrors Since Ages Ago™


TheoM
229 posts

Master Geek
+1 received by user: 55

ID Verified
Trusted

  #2235922 13-May-2019 12:56
Send private message

edc:

 

As I've stated many times, I do notice the difference post change. The ping is measured consistently from the router, in client (in game). It isn't wireless (I'm cabled), it isn't Windows, it isn't my hardware. Nothing has changed apart from the ISP's performance. [...]

 

 

Realtek or Atheros based NIC? They're onboard NICs for a reason. They're very cheap, and can definitely add lag. Complaining about 10ms is also futile in reality, your lag also depends on the tick rate of the server, which from what I've read, they run at a tick rate of 30hz, or about 33.3ms per tick. 10ms is a non issue if you're still waiting for the server to respond to your client.

 

Checked with a live copy of Ubuntu Linux? an Intel NIC? an ISP provided CPE like the HG659 or Fritz!Box?





Hi! I'm TheoM, but you know that already. I run Linux mirrors in NZ together with 2degrees. Like a mirror added? PM me!

 


 

https://theom.co.nz | https://theom.nz | https://mirrorlist.mirrors.theom.nz | Providing Free Mirrors Since Ages Ago™


 
 
 

Shop on-line at New World now for your groceries (affiliate link).

edc

edc

31 posts

Geek


  #2235927 13-May-2019 13:10
Send private message

TheoM:

 

edc:

 

As I've stated many times, I do notice the difference post change. The ping is measured consistently from the router, in client (in game). It isn't wireless (I'm cabled), it isn't Windows, it isn't my hardware. Nothing has changed apart from the ISP's performance. [...]

 

 

Realtek or Atheros based NIC? They're onboard NICs for a reason. They're very cheap, and can definitely add lag. Complaining about 10ms is also futile in reality, your lag also depends on the tick rate of the server, which from what I've read, they run at a tick rate of 30hz, or about 33.3ms per tick. 10ms is a non issue if you're still waiting for the server to respond to your client.

 

Checked with a live copy of Ubuntu Linux? an Intel NIC? an ISP provided CPE like the HG659 or Fritz!Box?

 

 

My setup is a Intel i350-T4-V2 NIC, CAT6a 550 MHz cables. The delay isn't on my infrastructure, it is outside of my network, on the ISP's, between their Auckland to Sydney hops. 

 

Add all the points at which lag occurs together, from the 8ms USB polling rate, to the refresh rate of the screen, to my reaction times, the last 10ms is the final straw that is bothersome. 

 

 

 

TheoM:

 

edc:

 

[...] I can see with the traceroutes that Auckland to Sydney should be about 23 ms. The pfsense setup is essentially out of the box, I only activated the 2 additional NICs. 

 

 

Go get yourself a cheap HG659/b and test using that. pf really isn't very useful for diagnosing anything. Anything from BSD's drivers for your NIC to the fact BSD's implementation of PPPoE is single threaded can make a huge difference.

 

While Auckland to Sydney can be 24ms, that is getting very close to the theoretical minimum RTT ICMP Echo time. A lot can influence that ping, from your internal routing configuration to the routers on the way (for example, since ICMP Echo is the lowest priority, if a router is busy enough it can elect to ignore the ping, or delay it until it has enough time to respond). 

 

As was said previously, it also depends on exactly which cable your traffic flows over.

 

 

The issue isn't on my infrastructure, that's what the traceroute and Wellington - Auckland pings confirm. The traceroutes were UDP and ICMP. 

 

The ISP's link between Auckland and Sydney is the issue, I will now find a ISP that uses the Southern Cross cable.


TheoM
229 posts

Master Geek
+1 received by user: 55

ID Verified
Trusted

  #2235971 13-May-2019 13:29
Send private message

edc:

 

My setup is a Intel i350-T4-V2 NIC, CAT6a 550 MHz cables. The delay isn't on my infrastructure, it is outside of my network, on the ISP's, between their Auckland to Sydney hops. 

 

Add all the points at which lag occurs together, from the 8ms USB polling rate, to the refresh rate of the screen, to my reaction times, the last 10ms is the final straw that is bothersome. 

 

[...]

 

The issue isn't on my infrastructure, that's what the traceroute and Wellington - Auckland pings confirm.

 

 

We're trying to help. Regardless of whether or not your pfSense appliance is capable of the connection, you do still need to do some tests with CPE, like the 659. Go back to the basics, and see if it improves before blaming your ISP.

 

As I've said previously the RTT can vary greatly, it all depends on what route your traffic takes. Without you performing the tests that I'm sure MR will get you to do, as would any other ISP, there's not much more we can do. 

 

I've ISP hopped many times between Spark, Voda, 2degrees, Voyager, MR, and a few other smaller ISPs. The routes are largely the same everywhere. ISP hopping isn't going to do much for your latency. It could be one of many things, or multiples of them that are causing you issues.





Hi! I'm TheoM, but you know that already. I run Linux mirrors in NZ together with 2degrees. Like a mirror added? PM me!

 


 

https://theom.co.nz | https://theom.nz | https://mirrorlist.mirrors.theom.nz | Providing Free Mirrors Since Ages Ago™


bagheera
545 posts

Ultimate Geek
+1 received by user: 190


  #2235974 13-May-2019 13:33
Send private message

if you got kb4482887 install on windows 10, then that is more likely your problem https://support.microsoft.com/en-us/help/4482887/windows-10-update-kb4482887 

 

 

 

"After installing KB4482887, users may notice graphics and mouse performance degradation with desktop gaming when playing certain games, such as Destiny 2."

 

 

 

 


noroad
1037 posts

Uber Geek
+1 received by user: 691

Trusted

  #2235983 13-May-2019 13:45
Send private message

TheoM:

 

I've ISP hopped many times between Spark, Voda, 2degrees, Voyager, MR, and a few other smaller ISPs. The routes are largely the same everywhere. ISP hopping isn't going to do much for your latency. It could be one of many things, or multiples of them that are causing you issues.

 

 

 

 

This being said, TGA and Hawaiki were not options to Australia until relatively recently, so things are quite possibly different. Some ISP's will now be using a longer/cheaper path that was not an option until fairly recently. The same applies to USA transit, not all path's are created equal as such.


1 | 2 | 3
View this topic in a long page with up to 500 replies per page Create new topic








Geekzone Live »

Try automatic live updates from Geekzone directly in your browser, without refreshing the page, with Geekzone Live now.



Are you subscribed to our RSS feed? You can download the latest headlines and summaries from our stories directly to your computer or smartphone by using a feed reader.