[BO3 Help] Entry Level BufferBloat – “Zennon’s First Rule of Gaming”

[dionysus]

That's one for the favourites bar, good stuff.

[Cajun72]

I've been testing the BB by gradually reducing the upload/download as advised until the green meter only gives me green readings with no spikes. This is on Reactive algo.

 

On the CC sliders this was around the 74 d/l and 56 u/l mark. However, although the reading for BB was 'A', the line quality was only 'B' and speed 'C'

 

Is there anything specific in the test which decides the rating of the line quality, and if so how can I improve? My paid for speeds are 76 d/l and 20 u/l.

 

I must admit after testing for about 20 minutes last night once I got the meter completely green and got an 'A' rating for BB I decided to go online and play cod.

 

So I guess what I am asking is, once I get the 'all green' reading on the meter, should I be happy with that or keep reducing the sliders in the hope I may get an A+ for BB and line quality?

 

Also if I can't get A+ all the way, am I correct in assuming it is much better to have a constant green meter with 'A' for BB and 'B' for line quality rather than 'A' and 'A' reading but the meter spikes red even once?

 

I will get more time at the weekend to do more tests so your expert advice is really appreciated !

[Dillinger]

HI Cajun - Let me see if I can help with some of the questions, but as always, you will need to try a few variants to be sure on your end.

 

The bufferbloat test should be performed while hard wired to the router.  There are big differences for my home personally if I do BB on the Wifi over a hard connection.  It takes way more slider to get into that magic green, and with the advanced settings we are fooling with, it rapidly degrades the house experience for the Mrs.

 

I have noticed that changing from Pre-Emptive to Reactive can have a significant difference in the hard wired reading for line quality.  I ran the test (4) times last night:

• Pre-Emptive,
• Reactive,
• Pre-Emptive 
• Reactive again

One setting (the Pre-Emptive) gave me A+, A+ for buffer bloat and line quality.  With the same settings, or even using more slider, I was never able to get above a 'B' for Line Quality on Reactive.

 

Now I am not ready to say that the difference in measured line quality by a third party sight that had nothing to do with either the Netduma or Black Ops 3 creations is "the ticket".  I do find it very interesting and that makes me think some further investigation is needed this weekend.

 

My limited take on this is such:  If we are obsessing with the connection ping ("I have a __ms ping and I am still losing every gunfight") then having a buffer bloat that can spike into the hundreds of milliseconds is bad.  Very bad.

 

With the gauge reading as close to zero (i.e. very little and ONLY green as possible), you are allowing for enough bandwidth to be preserved for gaming so that those spikes you can see in the gauge (when it goes to yellow or red) are eliminated.  In theory, this should mean your game packets are free to enter and leave the console without being overrun by "other" requests in the same line.

 

My theory is in line with your last statement:  If I can get an A+ for BB and for line quality, will my gaming experience be significantly better? Or will it be a marginal improvement rarely effecting much?

 

That leads me to the next round of tests.  OCD is bitch, let me tell you....LOL

 

Luckily the games in TDM have been extended to 100 kills and it's Double CrytpoKey weekend, so there will be plenty of targets.  LOL

[Cajun72]

Hi JD

 

Good to hear from you! Thanks for the write up. Yes I always test wired to PC. It sounds like not getting any yellow/red spikes is the key then I will try testing this weekend.

 

What I am finding so far is the yellow/red spikes are happening ONLY on the (blue) download test. Once it finishes and goes to the upload test (red) it always remains green even with slider for upload on 95% and barely goes above the 15ms mark so I'm hoping my upload slider setting is already good, I just need to work on the download being all green.

 

Running the test now on Saturday afternoon and with reactive by lowering my d/l slider to 69% I got A+ for BB but only got a C for line quality which is quite odd as the u/l test never went above 11ms!!

 

Very addictive this is turning out to be, almost as much as playing cod don't you know !  

[tonyvega]

does this all have to be on a enthernet connection....?? or can i test this wifi?

[tonyvega]

does this need to be test on a ethernet connection?

[Dillinger]

does this all have to be on a enthernet connection....?? or can i test this wifi?

 

You can test it with Wifi, but you will get much more accurate and better results if you plug in the hardline.

[Cajun72]

I can only get an 'F' on speed when managing to get A+ for BB and quality. Is this really bad and what effect will it have? Should I try to get a better speed rating which could lower the other ratings?

 

[Zennon]

I can only get an 'F' on speed when managing to get A+ for BB and quality. Is this really bad and what effect will it have? Should I try to get a better speed rating which could lower the other ratings?

 

The reason you are doing this is to lower your speed to leave headroom so the line does not saturate and cause "Buffer Bloat" while others are using the internet on your network which helps with gaming.

 

So no do not worry at all about speed

[UK_Wildcats_Fans]

I would not worry about the speed grade.  The first two are the most important.

[Cajun72]

Thanks, appreciated 

This is on pre-emptive with cc sliders both on 70 by the way.

[Zennon]

Thanks, appreciated 

This is on pre-emptive with cc sliders both on 70 by the way.

I prefer reactive but use what works best for you.

[Cajun72]

Up to the point this guide was written I also used reactive as I got good BB scores, mainly A.

 

After reading this guide I ran again and noticed the meter hit red a few times so I tried preemptive and got better BB scores with no red spikes.

[Zennon]

I run speed tests while running ping plotter so i get a more in depth look at whats happening and for some reason reactive make my line jitter by 1ms when flooding it where as preemtive can have  up to 20ms spikes.

[Dillinger]

Up to the point this guide was written I also used reactive as I got good BB scores, mainly A.

 

After reading this guide I ran again and noticed the meter hit red a few times so I tried preemptive and got better BB scores with no red spikes.

 

I noticed the same thing and ended up using what gave me the most green / least red for whatever time, day or system I was playing.  This has led to mostly positive experiences prior to dot6.

 

Now the bad hit detection games appear to be the anomaly and not the other way around....LOL

[Cajun72]

I run speed tests while running ping plotter so i get a more in depth look at whats happening and for some reason reactive make my line jitter by 1ms when flooding it where as preemtive can have  up to 20ms spikes.

Interesting...was you getting the red spikes on the meter as well for preemptive?

[Zennon]

I do not use the site only ping plotter as it is local to the machine I am testing which I prefer, after all ping is a there and back test.

[Cajun72]

Ok Thanks might give that a go later

[fuzzy clam]

I do not use the site only ping plotter as it is local to the machine I am testing which I prefer, after all ping is a there and back test.

So Z would you suggest me running ping plotter while I'm checking BB ?

[Zennon]

It's just a personal thing but running a ping plotter from a local machine test the ping there and back to the local for me is the correct way to test ping.

 

As ping is a destination and back travel time.

 

So doing this while testing the CC makes sense to me.

[Sir_smooch]

It's just a personal thing but running a ping plotter from a local machine test the ping there and back to the local for me is the correct way to test ping.

 

As ping is a destination and back travel time.

 

So doing this while testing the CC makes sense to me.

 

Ok so this is a bit duuuhhh, but just to clarify! You run a bb speed test at the same time as running ping plotter? Are you pinging 8.8.8.8 and speedtesting at DSL reports?

 

Oh and just wanted to say thanks to all you guys who are putting all the time and effort in to try and improve everyone's gaming experience. I know u don't need to do it, but really appreciate that you all are

[Dillinger]

Oh and just wanted to say thanks to all you guys who are putting all the time and effort in to try and improve everyone's gaming experience. I know u don't need to do it, but really appreciate that you all are

 

Thanks man, I know the guys all appreciate the feedback.  

 

We are all gamers too and we want to play good games, with good results and not rage either  Some of us, like Zennon, have a Yoda-esque quality to be able to share important information in such a fashion to spark interest & could be lost otherwise.  That gets the wheels spinning and OCCASIONALLY results in some useful info for the rest of you.  Granted the #Canary cage is a mess with 'tried this" and "tried that" threads, but the good stuff we being out....LOL  

[Zennon]

You run a bb speed test at the same time as running ping plotter? Are you pinging 8.8.8.8 and speedtesting at DSL reports?

 

 

Yes I run speed tests as I use pingplotter pinging 8.8.8.8

 

Speed testing to speedtest.net picking a close server to me to get the best throughput as dsl reports is US owned.

[Zennon]

Some of us, like Zennon, have a Yoda-esque quality

Included yourself you are.

[Sir_smooch]

Yes I run speed tests as I use pingplotter pinging 8.8.8.8

 

Speed testing to speedtest.net picking a close server to me to get the best throughput as dsl reports is US owned.

Cool, thanks Zen. I will deffo try this for sure. Is there anything in particular I should be looking for? Or will it be instantly apparent if something is wrong?