Warzone Lag - The usual Shoot first die first

[Netduma Fraser]

7 hours ago, Piplowe said:

I used 2 method to test,   i used the speed test at fast.com and the DumaOS 3.0 connection benchmark. 

As you can see from the images below, when my MTU was 1500 my loaded upload was 37ms,  reduced it to 700 and now it's 20ms and hit detection is good.

@Netduma Fraser Perhaps the devs could shed some light.  Since lowering the MTU I'm starting to play a lot more on the uk server and joining lower ping games.

Could the loaded upload ping be a factor when matchmaking rather than an idle or download loaded ping.  

So there is something to this. I don't think it will necessarily work for everyone, really depends on the quality of your connection in general. So a higher MTU is more efficient and processes fewer packets for the same amount of data compared to a lower MTU. However, higher MTU can cause network delay in some cases and if there are any errors which is likely on larger packets then the entire packet needs to be sent again and takes longer. To summarize, higher MTU will be best for most but some may find a decrease in latency using a lower MTU. I wouldn't recommend lowering it the way that has been done because it's still important not to have fragmentation so using this guide will help you ensure you arrive at a value that won't cause fragmentation. 

https://kb.netgear.com/19863/Ping-Test-to-determine-Optimal-MTU-Size-on-Router

[N3CR0]

I remember experimenting with MTU values back on Infinite Warfare.

I found overall it made everything worse.

Some gunfights I would just sponge bullets... I could see they were hitting me but doing almost no damage (MarkOfJ style lol)

But then more times than not the other player would be the sponge... and at the time that was rare but messing with my MTU made it happen several times a match.

Not saying it couldn't possibly work out well at all times for some peoples (we all have different connections afterall)... just stating my experience with it.

[Piplowe]

2 hours ago, Netduma Fraser said:

So there is something to this. I don't think it will necessarily work for everyone, really depends on the quality of your connection in general. So a higher MTU is more efficient and processes fewer packets for the same amount of data compared to a lower MTU. However, higher MTU can cause network delay in some cases and if there are any errors which is likely on larger packets then the entire packet needs to be sent again and takes longer. To summarize, higher MTU will be best for most but some may find a decrease in latency using a lower MTU. I wouldn't recommend lowering it the way that has been done because it's still important not to have fragmentation so using this guide will help you ensure you arrive at a value that won't cause fragmentation. 

https://kb.netgear.com/19863/Ping-Test-to-determine-Optimal-MTU-Size-on-Router

I agree with you there Fraser it wont work for everyone, just trying to work with what i've got.
My ISP (Sky) wont investigate further if I'm within their acceptable service limits.  They do not guarantee upload speeds so I'm assuming that also has a knock on affect to latency and prioritisation on their side.  

I've tested the link above before and had no joy, the only way I've managed to get some consistency is by bringing my upload loaded latency down via the MTU on netduma
Interestingly, if I just change the MTU on the playstation and not the netduma, I cannot join any games in warzone.  

[davemarsden132]

Thanks alot for sharing that Piplowe - Im going to have a try tonight and see if it works for me or not. Will be able to come back with the results but appreciate the time and effort into replying =] 

[Netduma Fraser]

19 minutes ago, Piplowe said:

I agree with you there Fraser it wont work for everyone, just trying to work with what i've got.
My ISP (Sky) wont investigate further if I'm within their acceptable service limits.  They do not guarantee upload speeds so I'm assuming that also has a knock on affect to latency and prioritisation on their side.  

I've tested the link above before and had no joy, the only way I've managed to get some consistency is by bringing my upload loaded latency down via the MTU on netduma
Interestingly, if I just change the MTU on the playstation and not the netduma, I cannot join any games in warzone.  

Likely because the Xbox will be sending packets at a lower MTU which conflicts with the higher value on the router and causes issues. Always best to do it on the router itself

[davemarsden132]

This is my before and after for MTU, 

At 1500 i get a peak of 40ms, at 680 i get a peak of 20ms..

 

Why does mine shoot up drastically at start and then calm down? Is that something in my settings or just something i cant change

[Netduma Fraser]

19 minutes ago, davemarsden132 said:

This is my before and after for MTU, 

At 1500 i get a peak of 40ms, at 680 i get a peak of 20ms..

Why does mine shoot up drastically at start and then calm down? Is that something in my settings or just something i cant change

We need to make improvements on the XR routers for the feature so that's why it spikes, very unlikely to actually be doing that

7 minutes ago, GBDJG said:

Would be nice if there were a app or program to start at 0 and determine the value.

Damn im lazy.😶

I think there are some but I haven't used them so couldn't attest to their quality

[Piplowe]

There is a web version, again may not be completely accurate, also make sure you test on a wired connection. 
 

http://www.letmecheck.it/mtu-test.php
 

I still get 1500 recommended, Odd that I get fragmented half way through. 

Sending 32 bytes to XX.193.11.9  <-  not fragmented

Sending 750 bytes to XX.193.11.9  <-  not fragmented

Sending 1125 bytes to XX.193.11.9  <-  not fragmented

Sending 1313 bytes to XX.193.11.9  <-  not fragmented

Sending 1407 bytes to XX.193.11.9  <-  not fragmented

Sending 1454 bytes to XX.193.11.9  <-  not fragmented

Sending 1478 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1466 bytes to XX.193.11.9  <-  not fragmented

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

Sending 1475 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1473 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

[Netduma Fraser]

45 minutes ago, Piplowe said:

There is a web version, again may not be completely accurate, also make sure you test on a wired connection. 
 

http://www.letmecheck.it/mtu-test.php
 

I still get 1500 recommended, Odd that I get fragmented half way through. 

Sending 32 bytes to XX.193.11.9  <-  not fragmented

Sending 750 bytes to XX.193.11.9  <-  not fragmented

Sending 1125 bytes to XX.193.11.9  <-  not fragmented

Sending 1313 bytes to XX.193.11.9  <-  not fragmented

Sending 1407 bytes to XX.193.11.9  <-  not fragmented

Sending 1454 bytes to XX.193.11.9  <-  not fragmented

Sending 1478 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1466 bytes to XX.193.11.9  <-  not fragmented

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

Sending 1475 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1473 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

It's just saying the maximum MTU you can input rather than what they recommend

[davemarsden132]

2 hours ago, Piplowe said:

There is a web version, again may not be completely accurate, also make sure you test on a wired connection. 
 

http://www.letmecheck.it/mtu-test.php
 

I still get 1500 recommended, Odd that I get fragmented half way through. 

Sending 32 bytes to XX.193.11.9  <-  not fragmented

Sending 750 bytes to XX.193.11.9  <-  not fragmented

Sending 1125 bytes to XX.193.11.9  <-  not fragmented

Sending 1313 bytes to XX.193.11.9  <-  not fragmented

Sending 1407 bytes to XX.193.11.9  <-  not fragmented

Sending 1454 bytes to XX.193.11.9  <-  not fragmented

Sending 1478 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1466 bytes to XX.193.11.9  <-  not fragmented

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

Sending 1475 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1473 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

PIPLOWE I COULD FUCKING KISS YOU!!!!

It worked for me ladies and gentlemen, full speeds, full ping.... changed MTU value to 680 which was good for me... here is the results! 

[L7NALMARI]

22 hours ago, Piplowe said:

I may have a fix for some !! 

I've had the same problem for years, low ping and poor hit detection, slow movement, 5 hit markers taking one plate in warzone etc. 
I've tried the normal routes to get better hit detection, VPN, Geo Filter to Play on servers further away, traffic priority, congestion control, throttling upload and download and even used an openwrt router to add ping delay.  I generally think you're fighting your ISP for a stable upload connection.  

Last week I was browsing the forum and saw that someone else changed their MTU to WAN setting to a low number and had better hit detection.  ( Fraser mentioned fragmentation can occur at low speeds)  So I thought I'd give it a go. 

As a side note I first noticed that my loaded ping upload/download where miles apart.  15ms download and 40ms upload, with no other devices on the network.

By lowering the MTU I was able to reduce the loaded upload ping from 40ms to around 25ms. I started with the default MTU of 1500 and lowered by 100 each time and testing the loaded ping on fast_dot_com. I settled with an MTU of 700 which is pretty much the lowest setting. 

I'm not sure why it works but I'm not complaining and my hit detection is night and day. 
I'm assuming it's something to do with either UDP packet loss or retransmission of UDP packets on high speed lines such as FTTC.

Give it a try, only change the MTU on the netduma and not your router, pc or console. 

 

Whats your set up ..do u connect your netduma router to other router with ppoe ? Or without it ? 

[L7NALMARI]

2 hours ago, Piplowe said:

There is a web version, again may not be completely accurate, also make sure you test on a wired connection. 
 

http://www.letmecheck.it/mtu-test.php
 

I still get 1500 recommended, Odd that I get fragmented half way through. 

Sending 32 bytes to XX.193.11.9  <-  not fragmented

Sending 750 bytes to XX.193.11.9  <-  not fragmented

Sending 1125 bytes to XX.193.11.9  <-  not fragmented

Sending 1313 bytes to XX.193.11.9  <-  not fragmented

Sending 1407 bytes to XX.193.11.9  <-  not fragmented

Sending 1454 bytes to XX.193.11.9  <-  not fragmented

Sending 1478 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1466 bytes to XX.193.11.9  <-  not fragmented

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

Sending 1475 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1473 bytes to XX.193.11.9  <-  FRAGMENTED!

Sending 1472 bytes to XX.193.11.9  <-  not fragmented

why i got this

[Piplowe]

You're welcome, it feels good to get that hit reg !!! 

I think it'll be something that needs constant tweaking though,  I just jumped on cold war and it was worse, so perhaps something slightly different in the netcode between the two games. 

 

 

1 hour ago, davemarsden132 said:

PIPLOWE I COULD FUCKING KISS YOU!!!!

It worked for me ladies and gentlemen, full speeds, full ping.... changed MTU value to 680 which was good for me... here is the results! 

 

[Infoseye]

This sounds too good to be true Adjusted MTU from 1500 to 680 ping unload reduced from 305 to between 9ms & 20ms I'm gonna play some games see what's like

 

 

[TODDzillaInLA]

i dont know if it will let you go that low i think 700 is the cut off and 1500 is the max

[Infoseye]

750 seems to be sweet spot for me and wow the game plays different, its much smoothed movement and responsiveness. Hit reg is crazy good.

[N3CR0]

Did some testing this morning, tried many different MTU values and found it made 0 difference to my ping on load.

Every single value I tried still gave me 80+ on load.

Just my two cents.

[Kevin.pal]

On 9/3/2021 at 5:19 PM, Seye said:

750 seems to be sweet spot for me and wow the game plays different, its much smoothed movement and responsiveness. Hit reg is crazy good.

Do you lower it to your console and netduma? My games don’t load with my Mary value at 750 on both units 

[zman1156]

14 hours ago, Swgohlegend said:

Do you lower it to your console and netduma? My games don’t load with my Mary value at 750 on both units 

no matter what always make changes on the router itself and not on the console otherwise that may create issues

[shiddy]

will this work only for warzone or also black ops cold war? what mtu settings do you guys use for CW?

 

[zman1156]

11 hours ago, shiddy said:

will this work only for warzone or also black ops cold war? what mtu settings do you guys use for CW?

 

i think most of the people here are talking about warzone. Since cold war is on a different engine and stuff mtu values will probably be different. Not sure but that is my guess. 

[TODDzillaInLA]

On 9/4/2021 at 5:15 AM, N3CR0 said:

Did some testing this morning, tried many different MTU values and found it made 0 difference to my ping on load.

Every single value I tried still gave me 80+ on load.

Just my two cents.

Depends on where you are changing your mtu value and if your playing on console or pc

On 9/7/2021 at 12:25 AM, Swgohlegend said:

Do you lower it to your console and netduma? My games don’t load with my Mary value at 750 on both units 

i dont know about playstation but on xbox you can not change your mtu setting

and from what i been reading i think alot of you messing with your mtu is more of a placebo effect

if you have a duma r2 and you change your mtu setting under wan / network / mtu it will not change your xbox mtu speed

[N3CR0]

5 hours ago, TODDzillaInLA said:

Depends on where you are changing your mtu value and if your playing on console or pc

i dont know about playstation but on xbox you can not change your mtu setting

and from what i been reading i think alot of you messing with your mtu is more of a placebo effect

if you have a duma r2 and you change your mtu setting under wan / network / mtu it will not change your xbox mtu speed

Not playing on anything, I don't play CoD anymore. I was testing on the Fast . com site same as a few others here.

I was changing the MTU on my main router.

[Piplowe]

As @Netduma Fraser mentioned changing the MTU may only work for a few people and may only work with . 
Changing the MTU seems to help if you have a poor upload connection, I'm not talking about upload speed, more about ISP network performance/throughput
When I watch others play, I see them lock on and melt players effortlessly. What I have now is still not close to that but it's better than what I had and certainly no placebo effect. 

My personal opinion is that if you have poor upload performance, UDP packets are getting delayed/out of order and the game servers are just dropping them as it's a time critical application. 

Quote

Time-sensitive applications often use UDP because dropping packets is preferable to waiting for packets delayed due to retransmission, which may not be an option in a real-time system

In COD I've had many gun fights where I'll start landing hits first on an opponent and continue to get hits till I die in what seems like a fraction of the time they land their first hit. On the kill cam (pinch of salt) I see their life bar almost drop to nothing before they land a shot and then they stop taking any more damage and kill me.   
My theory here is that the game is trying to see who fired first and who did the most damage in that time frame and if your packets are getting delayed/out of order on upload you are up against it. Smaller MTU size seems to narrow that gap for me in terms of assisting with my poor upload performance. 

Basically the games netcode is trying to handle all the variables in everyones connection in a round trip from their home to the game server and back, whilst trying to create a level playing field in real time. It must be challenging considering the differences in Home networks, ISP Hardware, ISP performance,  local network congestion, game server congestion etc. 
To put it into perspective of how quick it's all happening - If your are playing in a 25 ms ping lobby, that is 10 times quicker than the average humans reaction time. 
The nearest Cold war game server to me is 3ms yet I'll end up in a game on a server that's running at 40ms. 

Let's be honest, most of us are here because we had issues with gaming/network performance and bought a Netduma product to fix something, others seems to do just fine with their stock ISP router. Whether you're combatting bufferbloat, high ping, Qos, traffic prio, port forwarding, geo fencing or having to force further away game servers,  the netduma helps us manage a poor gaming experience.  

Give it up for the Netduma team !!!!!
Not only do they give us a product that allows us to tweek and test our connections, they also provide support and suggestions to things that are way out of their control. 

 

[Infoseye]

On 9/7/2021 at 8:25 AM, Swgohlegend said:

Do you lower it to your console and netduma? My games don’t load with my Mary value at 750 on both units 

Router