No FPS gaming possible

[donger]

I would actually give zero attention to bufferbloat because it gave my 1000/1000 fiber connection a C rating when i have 0 packet loss and absolute perfect performance. I ofc don't game when downloading with 100% speed because that would just be idiotic.

just for the sake of sharing data, that's how it looks like on literally prefect fiber 1000/1000 connection. I have zero issues in games and obviously i am not downloading p0rn when playing...

Great so it's just another useless "test" spewing out random numbers. I don't have pingspikes or anything like that when gaming... even when downloading games and playing at the same time. Just like some other BS website rated my connection C lmao

Judging by the post in which you state: I don't have pingspikes or anything like that when gaming... even when downloading games and playing at the same time.

Why then did you write earlier: I ofc don't game when downloading with 100% speed because that would just be idiotic

If you don't have any performance issues in games when using the channel at full speed, then why is it idiotic to do this?

There must be logic here, it’s a pity that I can’t understand it with my psychiatric deviations.

I usually don't do it but it happens sometimes when steam, EA app or another crap decides that it's a good time to update my game. Why would i risk it if i don't have to? Ur just spamming at this point

[1000WATT]

I usually don't do it but it happens sometimes when steam, EA app or another crap decides that it's a good time to update my game. Why would i risk it if i don't have to? Ur just spamming at this point

So you're accepting the risk of gameplay instability when the bandwidth is completely full. The test is completely useless, as during normal gaming there's no such load, and the risk of gameplay instability is minimal.

If this test could measure jitter under a real gaming load of, say, a few megabits per second, it might be able to detect the presence or absence of gaming issues. But it doesn't do that, and therefore is completely useless.
Correct me if anything is incorrect.

[LaggyTyp]

First of all: I haven't been able to find a solution to my problem, but I know it's the ISP's fault. It's the upload channel. It's not developed here (only one line) and is constantly overloaded. My download and upload pings are different, plus there's jitter in the upload.

But one thing isn't making sense to me!

I live on side A of the street.
My neighbor is on side B.
Our houses are maybe 30 meters apart.
I have cable internet (Docsis 3.0 + 3.1) from Vodafone.
He has VDSL from O2.
I play Battlefield 2042.
He plays The Finals.
If I complain about problems, he has them too, and vice versa.

How is this possible?

[1000WATT]

4.jpg (271.24 KiB) Viewed 956 times

If Hub 3 is overloaded, all houses will have problems.
Or do you think that everyone has a separate cable going directly to the backbone provider's server?

you're just joking

[Crozaa]

Been following this for a while, I’m on a FTTP connection in the UK with BT(900/100) and suffer almost the exact same symptoms. Latency jitter on unloaded, download, upload which varies at different times of day. I’ve seen upload spikes as high as 600-700ms, and any time I see these figures across various speedtesters, hitreg, recoil and smoothness is terrible. I know bufferbloat tests are synthetic and quite unrealistic in terms of the environment you’d see whilst playing cs2, but they still give good data for measuring certain times of day, which can be overlayed with both subjective feeling in game, and in cs2, frame times themself can give an indication.

I did recently get a Flint2 router as I’ve went for a new 2.5 symmetrical contract, and whilst configuring it I have managed to reduce most of the standard buffer lost from isp router, where I can consistently get 0-2ms latency differences on all 3, by applying an aggressive SQM setup. This however never fully gets rid of these crazy upload spikes, “rush hour” times I struggle to get through a test (multiple sites) on any device without seeing a reading of 150-200+ms.

I fully believe it’s all isp side, either with congestion or aggressive rate control. I will be able to test on Monday when I get new line installed.

[LaggyTyp]

There are only two ISPs here. The others are just renting.

Telekom = VDSL
Vodafone = Cable

Any idea how to test whether the hub is overloaded? Tracert doesn't provide anything useful.

[1000WATT]

Any idea how to test whether the hub is overloaded? Tracert doesn't provide anything useful.

There is no such possibility from the client's side.

[themagic]

@LaggyTyp

I have idea but don't sure if this will help you but try get on Amazon a VPS and with VPN and connect and try play.

Or try to get dedicated IP...

tried this way already ?

Are you having right now a Private IP or Public or ones which changes ip dynamic from your ISP ?


This would be interesting to know...


edit:

Look...the days where i had normal online gaming here and where I was happy with 1und1... but i have still no fuckk idea what happened since that and why everything become this bad here and where dont even provider change helped me or even mobile internet and today.

But i remember that i had a VPS server with 1und1 back then in this days but i think they dont offer this right now today and its now IONES different and seperat i think but where you can get a VPS too..

but back then i know for sure that i had a IP which was PUBLIC i think and which was dynamic and changed every time when i restart my Fritzbox...


Right now with Vodafone i have Private IP and Mobile from O2 too...


Like maybe that our Issues ? Just a idea...



Another Idea and thing which you can try: download WARP for windows and disable IPv4 in network card settings and use IPv6 only and connect to WARP...

https://developers.cloudflare.com/cloud ... load-warp/

Or try use Hideme VPN and same way and with IPv6 only...
https://hide.me/de/

you can use it 7 days for free and even port forwarding exist as option there.

you notice any difference ? pls tell when try this...