Using ONT Modem with ASUS Router advice

[KevyMatts]

Hi,

So my ONT Fibre modem in my router wireless log i can see that its using googles 8.8.8.8, so on my router should i use the same or is it okay to use cloudflare 1.1.1.1?

Also I have 300/100 FTTH package, should I use CAKE? currently without it i have Grade B. Using CAKE I can get A+ with 259.26/87.57





Trying to optimize my connection for stabilty i have notice in speed tests have increased Ping to 25ms - 30ms and Jitter at times can spike 15ms -17ms with Latency from 14ms to 50ms

I have a wired connection to gaming PC, then in house have couple of firesticks, Smart TV, mobiles, and a powerline adapter wired into router with wifi extender

 

[L&LD]

What router do you have? What firmware does it support (stock/RMerlin)?

Do you see a noticeable difference in network performance when browsing? Or, are just the Grades better?

You can use any DNS you want.

 

[KevyMatts]

Hi,

RT-AX68U was running merlin 386.4 was fine, then upgraded yesterday to 386.5 so far be rock stable, have diversion set up also.. Browsing i dont seem to notice any difference, just that grades better but if i was playing a game of COD my mother 4K Firestick might buffer but that is running of the powerline adapter/extender.

Oh okay i'll just stick what DNS i have set.

I havent no gaming today, but i have set up Broadband Quality Monitor and just checked it;

 

[EternalStudent07]

That packet loss is worrying. I assume it means no connection at all.

QoS just tries to allocate network resources across devices differently than they'd get otherwise. Normally certain uses will drown out others. QoS can keep some reserve capacity for the other people to use a bit of, so your speed tests won't show your full capacity. But it should make multiple people using the service a better experience for everyone.

I'd hope CoD wouldn't need much total bandwidth after updates happen, just that it would want very short latency. There are client side (OS network adapter) settings that claim to improve that, but I can't say how much they help. Often something can lower resource use or should give you better latency.

How much of your connection does the 4K Firestick use? I'm guessing a lot at first while it tries to buffer up some of the show initially.

 

[Paliv]

Gaming doesn’t use much up or down link bandwidth. It certainly should not be saturating a 300/100 connection to the point you have buffer bloat and need QoS. Something else must be going on.

 

[KevyMatts]

I think the Broadband quality monitor graphs showing 100% is due to he firewall in your hub is set to reply to ICMP pings. So i have read online.

Anything i can do to test or check my lines?

 

[EternalStudent07]

I've never used that tool before, but found this page... https://www.thinkbroadband.com/faq/broadband-quality-monitor

"

Why is my graph all red / show 100% packetloss?​

If your graph is solid red like this example:

BQM graph with ICMP blocked

Then double check that your registered IP address is correct, and that your broadband router that has been assigned this IP address. Check the router is configured to respond to ping requests on its WAN/Internet port. The option is sometimes referred to as 'reply to ICMP requests'.


We maintain a list of some routers with guides on how to configure them to work with the Broadband Quality Monitor in our Router Configuration Guides.


Our traceroute tool at www.thinkbroadband.com/tools/traceroute is an easy way of checking whether your router is responding, rather than waiting for the graph to update.


The latest version of the Virgin Media Superhub firmware (R36) hides the setting under Advanced settings > Tools > Ping - 'WAN Ping Respond', 'Respond to ICMP echo requests sent to WAN IP'
"

Yes, if you changed a setting at 10pm to block pings, or you got a new IP address that you didn't register with them then your graph would make sense. I don't know what you were trying to offer with that though.

It's hard to watch home network traffic unless your router has a built in feature like a bandwidth usage meter, and maybe a log to understand it's decisions and actions from. If you're using all your connection then response times increase while they wait their turn.

Typically I try to test each part separately is working well. Then check for issues when you use them together at the same time. Do the same thing and see if you get the same response. That'll hopefully point at where to adjust things, or where to dig deeper for a change to try.

I'd love to be wrong though. There is a reason network specialists get paid big money in the business world. It takes specialized equipment and knowledge to do well.