What's new

FlexQoS FlexQoS 1.3.2 - Flexible QoS Enhancement Script for Adaptive QoS

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Status
Not open for further replies.
Hey guys,

I have machines on my network making OpenVPN and Wireguard connections via UDP.

How do I properly shape those using FlexQoS? Is it enough to add a rule with the VPN destination port to shape its traffic or is there something else I need to do? Is this even possible?

Thank you!
If the encrypted tunnel is established between a VPN client on the LAN client and the remote VPN server, then there probably isn’t a lot that can be done to create fairness within the tunnel. What you can observe in the Tracked Connections list is all you can work with via the iptables rules.
 
If the encrypted tunnel is established between a VPN client on the LAN client and the remote VPN server, then there probably isn’t a lot that can be done to create fairness within the tunnel. What you can observe in the Tracked Connections list is all you can work with via the iptables rules.
Thanks! I don't want fairness within the tunnel, but fairness "outside" of the tunnel, e.g.: shape tunnel bandwidth in relation to other open connections going through my non-tunneled-uplink.

I might be confusing things, sorry if I explained it badly.
 
Thanks! I don't want fairness within the tunnel, but fairness "outside" of the tunnel, e.g.: shape tunnel bandwidth in relation to other open connections going through my non-tunneled-uplink.

I might be confusing things, sorry if I explained it badly.
In that case, it seems you have the right idea. Shape by port, or perhaps port+IP if the IP won't change much. You'll keep the tunnel's bandwidth constrained to whatever category you choose. You can verify afterwards that it is working by starting it up and downloading something through it, while checking the FlexQOS graphs from another computer.
 
Hi all, I'm having a strange issue in that when using flex, live video streams and Alexa music breaks up for a second every so often. Doesn't happen with Flex turned off.
 
Hi, I have recently upgraded to the firmware 386.4_alpha3 on my RT-AC88U and immediatly noticed that both download and upload graph almost stopped working and only register a couple of KB, I think the entire thing just stopped working because it used to help a lot to prevent lag spikes on my online gaming and just started to see them more frecuently. I'm on the latest FlexQoS version also did a hard reset to the router with no luck.
 

Attachments

  • graph.png
    graph.png
    32.8 KB · Views: 211
Last edited:
Hi, I have recently upgraded to the firmware 386.4_alpha3 on my RT-AC88U and immediatly noticed that both download and upload graph almost stopped working and only register a couple of KB, I think the entire thing just stopped working because it used to help a lot to prevent lag spikes on my online gaming and just started to see them more frecuently. I'm on the latest FlexQoS version also did a hard reset to the router with no luck.
Try to uninstall FlexQoS and see if the same problems occur with just Adaptive QoS on the Classification tab (i.e. no cumulative data in the pie charts or Total column). If so, report it in the 386.4 alpha thread.
 
Are you still using this setup on the screenshots? I've setup this way recently to include the bandwidth percentages and all is working pretty well. I was contemplating in moving streaming a little higher but I'm using your setup for now. I appreciated.
 
Last edited:
What uses net control? So we dont need to put 100% on net control like for gaming?
 
There is something wrong.
—File transfer category—
Start with a game download from Xbox (all band for this)
Start with a LG firmware download (all band for Xbox)

Pause Xbox download.
All band to LG firmware download.

Restart Xbox download ,all band to Xbox.

♂️
Confirmed. I was able to reproduce something like this. When an XBox game download starts, it spawns around 8 connections. And then everything dies and slows to a crawl. Those connections plus a bunch of others - they end up in Web Surfing. Yet in FlexQOS you have tons of bandwidth remaining, and in the WAN bandwidth monitor, the connection is completely redline. It suggests that somehow XBox game downloads bypass QOS and hog a ton of bandwidth. (UDP spam maybe? Please, never use UDP for downloads... no throttle mechanism. TCP only.)

1639299115844.png


This is a "25mbit" (30mbit, 37mbit DSL sync) connection. Screenshots taken at the same time. One shows 8mbit in use. The other is completely maxed out. During the spike websites can take minutes to load.

1639299156738.png


Once the game download finishes, everything goes responsive again.

1639299505891.png


Now I'm starting to wonder if Flow Cache, etc, is possibly the culprit? I need to do further testing.

1639299701237.png


I'll disable it with fc disable over Putty and then do some further testing when there's another XBox game download to do...
 
Are you still using this setup on the screenshots? I've setup this way recently to include the bandwidth percentages and all is working pretty well. I was contemplating in moving streaming a little higher but I'm using your setup for now. I appreciated.
Yep, still laid out exactly like that where I have 180mbit. It's working very well.

I have discovered an issue with XBox's at a second location, so I may need to make a few adjustments once I get that one wrestled down. :)


Net Control packets hardly use bandwidth. I have hardly ever seen them over 2%, even with massive torrenting, etc. - I set the maximum a bit lower because they are so high priority, if the wrong thing ended up in that class. I wouldn't want it to strangle everything else.
 
what do you think is the best category priority list for max "gaming" and no interfere, ping spikes with many other devices who surfs and watches youtube or just google around, includes pcs, Tvs and phones?
 

Attachments

  • Unbenannt.PNG
    Unbenannt.PNG
    36.8 KB · Views: 233
I usually put Gaming near the top, with reduced max bandwidth to stop a mis-categorisation playing havoc with things. (For example, a game download ending up in that category. Downloads can hog 100% of a connection.) Keep in mind that many games do NOT require much bandwidth to play. I have seen a lot of games that work on 256kbit-1.5mbit just fine. If you can grant them even a dedicated 20mbit, you're probably good for 99% of games. And a dedicated 5mbit will cover several users for > 90% of games. Some of the most popular games out there on Steam run fine on 1mbit or less. (It needs to be stable, though, so being categorized near the end would not work well.)

My current problem is that an XBoxOne is escaping FlexQOS entirely and hogging the entire connection for downloads. Under Traffic Analyzer -> Statistic, I can see the XBOXONE device chugging bandwidth at an insane pace, yet FlexQOS insists only 8 out of 30mbit is in use (in "Web Surfing" - mostly the XBox One)

I seem to remember a thread mentioning some bugs related to the RT-AX86U and its gaming port, and stuff bypassing qdiscs. I skimmed it a while back, but something in the back of my mind is telling me to go find it, as there might be useful info there.
 
Yep, still laid out exactly like that where I have 180mbit. It's working very well.

I have discovered an issue with XBox's at a second location, so I may need to make a few adjustments once I get that one wrestled down. :)


Net Control packets hardly use bandwidth. I have hardly ever seen them over 2%, even with massive torrenting, etc. - I set the maximum a bit lower because they are so high priority, if the wrong thing ended up in that class. I wouldn't want it to strangle everything else.
Please post your adjustments once you have fixed the issue. I appreciate it.
 
Hi, dave14305. I am using AX86U with Asuswrt-Merlin 386.3.2 and FlexQoS 1.3.0. I did not add or modify the iptables rules, nor did I manually disable Runner and Flow Cache. Now classification/flex show the bandwidth correctly, and I get A in bufferbloat test.
FlexQOS seems to work fine. So do I need to manually disable Runner and Flow Cache?I'm not sure if Runner and Flow Cache on AX86U will affect FlexQoS in some other way.
 
Hi, dave14305. I am using AX86U with Asuswrt-Merlin 386.3.2 and FlexQoS 1.3.0. I did not add or modify the iptables rules, nor did I manually disable Runner and Flow Cache. Now classification/flex show the bandwidth correctly, and I get A in bufferbloat test.
FlexQOS seems to work fine. So do I need to manually disable Runner and Flow Cache?I'm not sure if Runner and Flow Cache on AX86U will affect FlexQoS in some other way.
If everything is working correctly, no need to do anything.
 
Status
Not open for further replies.

Latest threads

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top