Netduma Complete video guide

[VengeanceStriker]

Hey everyone VengeanceStriker here. Has anyone seen a complete netduma video out there on the net? I ask because I'm a hardcore online gamer and am getting increased requests on the router. Most know what I'm running the netduma but are intimidated by the complexity which I understand for someone that knows nothing about networking. From what I've seen out there are videos that cover one topic. I'm talking about doing a full breakdown simplified including on the fly adjustments, tweaking and troubleshooting. I just don't want to reinvent the wheel if there is already one out there.

[BIG__DOG]

check out sim's channel!

 

https://www.youtube.com/user/simjc74/videos

[RADDY1993]

Hey everyone VengeanceStriker here. Has anyone seen a complete netduma video out there on the net? I ask because I'm a hardcore online gamer and am getting increased requests on the router. Most know what I'm running the netduma but are intimidated by the complexity which I understand for someone that knows nothing about networking. From what I've seen out there are videos that cover one topic. I'm talking about doing a full breakdown simplified including on the fly adjustments, tweaking and troubleshooting. I just don't want to reinvent the wheel if there is already one out there.

As the dog said Sims channel is great for Netduma videos.

[ColonicBoom]

Sim's got some good breakdowns and is quite funny too. But I'd give the fine-tuning using 'dsl reports' website test a miss, that test is inconsistent and not even close to as useful as people think it is.

[Madman]

I agree Sims channel is awesome, I've made a couple of videos of going through settings as well:

NETDUMA MENU: https://youtu.be/n__xWi0hp1A

OPEN NAT TYPE: https://youtu.be/esu9Ib-vYZM

[VengeanceStriker]

I've checked him out. I've had mine for a year and have mastered it. I disagree with him on the mtu. The mtu setting will change depending on speed. I was running 25up/down and using preemptive, now I'm running 125up/down dedicated fiber using reactive. Hard set the mtu on the PS4 pro for the maximum number which is 2 before the number where packets are fragmenting. Dropped antiflood to 50/50. Avg Ms ping is mid 20s. I know this will be different for everyone. I also manipulate the location for my friends that live in a certain area. Ill optimize myself for that state and save it in a profile. Then pull hosts in their area, start a lobby with them in it and all will have good connections to the host. Obviously these are people that aren't running netduma. Another thing I do is pull the hosts ip using Cain & Abel. Geo locate those ips and run dslreports with selecting servers as close as possible to that ip. With all that said and done my buffernbloat avgs 1ms 58/58 up/down mid 20s avg ping just about anywhere.

[ColonicBoom]

The mtu setting will change depending on speed.

 

mid 20s avg ping just about anywhere.

 

I don't think this is correct.

If your MTU has needed to be changed it's because you changed ISP/line/equipment maybe(?), not the bandwidth itself.

You MTU is either set correctly or it's not.

Don't mess about with MTU unless you have the wrong number set, anyone that wants to find out should Google how to check it using Command Prompt / Terminal on Mac.

 

Anywhere? This is not going to be correct either. It's not going to be the same to London as it would be to New York, Sydney, Japan.

[RL317]

Just make sure you ignore the "lower your speeds so games apply the lag comp algorithm differently to your connection" suggestion lol. Anti-lag has nothing to do with bandwidth; it's the game's ability to "rewind time" by looking at snap shots of events according to real time (ie server or host perspective) and comparing them to what happened on the client's end in order to determine who fired first or landed the kill shot first. It's purely about latency.

 

I also once watched a video of his where he set out to prioritise bufferbloat but ignored the fact he was introducing packet loss by dropping his bandwidth to a certain point. I should add that device prioritisation seems to cause packet loss too if share excess is unticked, so I wouldn't follow any advice suggesting you use that at all, never mind doing so to reduce bandwidth for a "lag comp fix".

[VengeanceStriker]

Yes. Line is all fiber now. I do this for a living btw. Before it was 1460 and after the change, I ran the test and it's now 1472. I have 0 issues. When I pull the trigger I hit immediately. Ethernet cut to length as short as possible, running 4k monitor 5ms refresh rate at 60hz. This mtu thing is such a debate. If you set the do not fragment flag on a pc, when you use icmp without setting a size it will send max transmission at 1500 and will tell you that it has to retransmit the rest of the data. So.....if I send 1500 and the port config can only take 1472 then I'm retransmitting 28 bytes which introduces some sort of lag. Setting the mtu too low will also introduce lag because it's not sending enough. I will run Wireshark and sniff the traffic from the ps4 pro. I don't think the PS4 pro set to automatic means that it will automatically sense what size to send for mtu. Thoughts?

[VengeanceStriker]

Here is how I setup. Netduma in DMZ, port forward or use UPNP for playstation to get the nat type open. PS4 Pro add to playstation network in hyper-lane, and add PS4 pro to playstation network via Host Filtering. In device prioritisation set your computer out to where all other devices are 1. In my case, I stretch it to 85 upload and download with "share excess" ticket for both. I have 125 up and down. Set to "Reactive", Download and Upload cap set at 100%. Deep packet inspection "OFF". On the computer run a ping to test with the "do not fragment" flag set. Once you find the number that has no packet loss, jot that number down. go to dslreports and pick a region. Im in the US so I'll pick whatever state. BO3 and lots of games run on gameservers.com server farm if I'm not mistaken. I run 3 tests moving 5% down on upload download cap. Each time, looking at the full results of the test for minimal bufferbloat and ping. I stop moving down when the tests get worse. A+ Quality, A+ Bufferbloat and A+ speed is what I achieved in all states where I normally play people. Go to the ps4 and manually enter the mtu number that doesn't fragment. Reset distribution and then pull your ps4 or whatever gaming device out like I did my computer in my case, I pull my pro to 85. Now go HAM! I also, have had to adjust on the fly and add more bandwidth due to players with slower connections dragging the lobby down so I can pull host. I haven't got into that yet about the lobbies but I know I've remedied those 3 bar situations that others get when another person that's slower enters the lobby.

[RL317]

1472 is the value without the header or whatever; this is what cable and fibre will show on a command prompt ping test, but 1500 is the max when added on. For xDSL that's 1464 but in reality you'd use 1492.

[VengeanceStriker]

1472 is the value without the header or whatever; this is what cable and fibre will show on a command prompt ping test, but 1500 is the max when added on. For xDSL that's 1464 but in reality you'd use 1492.

Correct without the header. Now that I remember my max before fragment was 1444 so I added the 28 to get 1472 where it's set now. I mean, if there's proof that the auto sensing works that's great but there are no port flapping issues so I know it's the most optimal setting.

[VengeanceStriker]

As far as proof of results, this was the first game I played after the upgrade and readjusting the netduma. The mothership is my #1 scorestreak and the highest streak on the game. I clear lobbies like this one shown below.  So I have to be at the most optimal now. 

 

https://www.dropbox.com/s/655org68w60lzs5/Vengeance%20Game.JPG?dl=0

 

 

[bagsta69]

I disagree with a fair bit of what Sim suggests but hey each to their own, I would probably hold off on putting a lot of work into a video for the current Duma OS as the new one is not that far away and I think a lot of the confusing stuff for people less tech savvy will be sorted out. 

If you want to fire on ahead Im sure people would be most grateful but if its a lot of effort I might hang fire. 

[ColonicBoom]

In device prioritisation set your computer out to where all other devices are 1. In my case, I stretch it to 85 upload and download with "share excess" ticket for both.

 

go to dslreports and pick a region. I run 3 tests moving 5% down on upload download cap. Each time, looking at the full results of the test for minimal bufferbloat and ping. I stop moving down when the tests get worse. A+ Quality, A+ Bufferbloat and A+ speed is what I achieved in all states where I normally play people.

 

I also, have had to adjust on the fly and add more bandwidth due to players with slower connections dragging the lobby down so I can pull host. I haven't got into that yet about the lobbies but I know I've remedied those 3 bar situations that others get when another person that's slower enters the lobby.

 

 

DSL reports is at best pretty inaccurate. With a line that was unusable (for gaming) I could achieve A+ across the board.

In my opinion it is a worthless test as far as optimising your connection goes. 

 

I'm not sure what you're saying about adding more bandwidth, I understand if you're talking about games like MW3 where you could manipulate host by changing your bandwidth but where is the benefit when you're all on dedicated servers (BO3 / IW / BF1 / OW)?

You do this for a living so you know that having more bandwidth does not equal less lag, games don't use much bandwidth at all. Most player lag is caused by a poor line or poor QoS.

 

I'm just asking because I think this might confuse people and cause them to mess around with 'the wrong settings' unnecessarily.

[fuzzy clam]

Agreed Boom... DSL reports are overrated,Pingplotter is so much better on a flooded connection.

 

And the last I knew there weren't a lot of P2P games as Boom said, if your playing a newer game you're most likely playing on a dedi.

 

Which would mean you have no control over the "3 bar" or slower connections in your lobby.We can only control our own connection and optimism our own network,not much you can do about playing in a lobby with someone who's stealing his neighbors WiFi and gaming.

 

I like playing the older titles and if I can pull host at least I know the connection is solid for myself and the rest of the lobby.

[RL317]

I like playing the older titles and if I can pull host at least I know the connection is solid for myself and the rest of the lobby.

Do you know what's required to actually pull host? Do you need a bare minimum of xMb upload bandwidth or something? I can never find a straight answer online lol

[fuzzy clam]

Do you know what's required to actually pull host? Do you need a bare minimum of xMb upload bandwidth or something? I can never find a straight answer online lol

On Ghosts on PS4,there seems to be no particular reason why you get host.My ISP speeds are below/normal 60/5 and it takes a couple of games to get it but I noticed one I got it,I had it until I left the lobby.

 

Wish I had a magic formula to share with you but I can't.All I can say is try playing Ghosts and see what happens.

 

And I don't throttle my connection,I set my CC at 70/70 and call it good.But always get host when I play but couldn't tell you why,I would like to say solid connection but surely there are people playing with faster and more stable connections than mine.

[RL317]

I haven't played Ghosts for a few months but I never pulled host there. I got it once in four player hosted games out of 3000+ games of BO3, and got it quite a lot in 2015 playing BO1.

 

What would the Netduma show if you had host, or would it remove the ping graph? I've had all circles on the map bulge like I was host before, only to be transferred to a server in IW, but I know it picks lobby hosts before moving on to a server. I just don't know what it takes to be host. They must have a f*cked up way of choosing who's closest to the server and provides the best, most stable experience because it's always given to some jittery WiFi player 1000 miles from the server lol

 

Edit: I found out that in CoD4, there'd be a "bandwidth race" to determine who got host, and in later titles it would be repeated when host migrations were necessary. In BO1 and onwards this would apparently occur when 'Connecting to Online Services'.

 

I wonder if that remains the case for picking lobby hosts in IW. Either way surely 10Mb upload should be enough to host a game right? On an incredibly stable line too... ¯\_(ツ)_/¯

[fuzzy clam]

I haven't played Ghosts for a few months but I never pulled host there. I got it once in four player hosted games out of 3000+ games of BO3, and got it quite a lot in 2015 playing BO1.

 

What would the Netduma show if you had host, or would it remove the ping graph? I've had all circles on the map bulge like I was host before, only to be transferred to a server in IW, but I know it picks lobby hosts before moving on to a server. I just don't know what it takes to be host. They must have a f*cked up way of choosing who's closest to the server and provides the best, most stable experience because it's always given to some jittery WiFi player 1000 miles from the server lol

Yeah I'm with you on how they choose who's host,I'm sure it's a extremely complex process...like some throwing a dart at a dartboard blindfolded and the closest person to the dart is host.

 

If you pull host the Duma would have all the same size circles, in my experience anyway but we should all have the same results with that part of it.I didn't pay much attention to it at first and then starting bailing on lobbies in the middle of games as the "host migration" always came up.And in my case again this happens overtime I play the game, but with Ghosts only I'm guessing it would work the same on MW3,BO2 anything that's not on Dedi's.