FlexQoS FlexQoS - Asus RTAC86U - Huge UPLOAD BufferBloat

[mikefrommanchester]

Hi

I have setup a new Asus RTAC86U today.

I have successfully* installed Diversion and FlexQoS on Merlin 386.1_2

My connection is stable at Down 55Mbps and Up 16Mbps.

I set Asus QOS --> Adaptive QoS ---> Manual Download 50Mbps and Upload 12Mbps (have tried other parameters but all the same).

When testing my BufferBloat at http://www.dslreports.com/speedtest I am getting huge Upload Bufferbloat at over 100ms.

My Download BufferBloat is perfect at 0ms.

Without QoS on I am getting less BufferBloat on my Upload - approx 50ms. But obviously I do not want any BufferBloat.

On my previous router with LuCi I had configured Cake successfully and BufferBloat was far better.

Nothing else other than my PC is connected to my network currently.



My FlexQoS debug log is below - but I'm at a loss of how to resolve this.

If anyone could possibly shed light on this or point me in the right direction I would really appreciate it.

Many thanks



* Possibly completely messed it up

Spoiler: FlexQoS Debug

FlexQoS v1.2.1 released 2021-02-12

Debug date    : 2021-02-21 13:56:02+0000
Router Model  : RT-AC86U
Firmware Ver  : 386.1_2
DPI/Sig Ver   : 2.0.1 / 2.218
WAN iface     : ppp0
tc WAN iface  : eth0
IPv6          : disabled
Undf Prio     : 2
Down Band     : 51200
Up Band       : 12288
***********
Net Control   : 1:10
Work-From-Home: 1:13
Gaming        : 1:11
Others        : 1:17
Web Surfing   : 1:15
Streaming     : 1:12
File Downloads: 1:16
Game Downloads: 1:14
***********
Downrates     :    2560,    7680,   15360,   10240,    2560,    5120,    2560,                                                                                5120
Downceils     :   51200,   51200,   51200,   51200,   51200,   51200,   51200,                                                                               51200
Uprates       :     614,    1843,    1228,    2457,     614,    1228,     614,                                                                                3686
Upceils       :   12288,   12288,   12288,   12288,   12288,   12288,   12288,                                                                               12288
***********
iptables settings: <>>udp>>500,4500>>3<>>udp>16384:16415>>>3<>>tcp>>119,563>>5<>                                                                            >tcp>>80,443>08****>7
-o br0 -p udp -m multiport --sports 500,4500 -j MARK --set-mark 0x8006ffff/0xc03                                                                            fffff
-o ppp0 -p udp -m multiport --dports 500,4500 -j MARK --set-mark 0x4006ffff/0xc0                                                                            3fffff
-o br0 -p udp -m multiport --dports 16384:16415 -j MARK --set-mark 0x8006ffff/0x                                                                            c03fffff
-o ppp0 -p udp -m multiport --sports 16384:16415 -j MARK --set-mark 0x4006ffff/0                                                                            xc03fffff
-o br0 -p tcp -m multiport --sports 119,563 -j MARK --set-mark 0x8003ffff/0xc03f                                                                            ffff
-o ppp0 -p tcp -m multiport --dports 119,563 -j MARK --set-mark 0x4003ffff/0xc03                                                                            fffff
-o br0 -p tcp -m multiport --sports 80,443 -m mark --mark 0x80080000/0xc03f0000                                                                             -j MARK --set-mark 0x803fffff/0xc03fffff
-o ppp0 -p tcp -m multiport --dports 80,443 -m mark --mark 0x40080000/0xc03f0000                                                                             -j MARK --set-mark 0x403fffff/0xc03fffff
***********
appdb rules: <000000>6<00006B>6<0D0007>5<0D0086>5<0D00A0>5<12003F>4<13****>4<14*                                                                            ***>4
filter change dev br0 prio 2 protocol all handle 828::800 u32 flowid 1:17
filter change dev eth0 prio 2 protocol all handle 828::800 u32 flowid 1:17
filter add dev br0 protocol all prio 2 u32 match mark 0x8000006B 0xc03fffff flow                                                                            id 1:17
filter add dev eth0 protocol all prio 2 u32 match mark 0x4000006B 0xc03fffff flo                                                                            wid 1:17
filter add dev br0 protocol all prio 15 u32 match mark 0x800D0007 0xc03fffff flo                                                                            wid 1:16
filter add dev eth0 protocol all prio 15 u32 match mark 0x400D0007 0xc03fffff fl                                                                            owid 1:16
filter add dev br0 protocol all prio 15 u32 match mark 0x800D0086 0xc03fffff flo                                                                            wid 1:16
filter add dev eth0 protocol all prio 15 u32 match mark 0x400D0086 0xc03fffff fl                                                                            owid 1:16
filter add dev br0 protocol all prio 15 u32 match mark 0x800D00A0 0xc03fffff flo                                                                            wid 1:16
filter add dev eth0 protocol all prio 15 u32 match mark 0x400D00A0 0xc03fffff fl                                                                            owid 1:16
filter add dev br0 protocol all prio 20 u32 match mark 0x8012003F 0xc03fffff flo                                                                            wid 1:15
filter add dev eth0 protocol all prio 20 u32 match mark 0x4012003F 0xc03fffff fl                                                                            owid 1:15
filter change dev br0 prio 22 protocol all handle 802::800 u32 flowid 1:15
filter change dev eth0 prio 22 protocol all handle 802::800 u32 flowid 1:15
filter change dev br0 prio 23 protocol all handle 804::800 u32 flowid 1:15
filter change dev eth0 prio 23 protocol all handle 804::800 u32 flowid 1:15
class change dev br0 parent 1:1 classid 1:10 htb  prio 0 rate 2560Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 32000
class change dev eth0 parent 1:1 classid 1:10 htb  prio 0 rate 614Kbit ceil 1228                                                                            8Kbit burst 3200b cburst 14400b quantum 7675
class change dev br0 parent 1:1 classid 1:11 htb  prio 1 rate 7680Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 96000
class change dev eth0 parent 1:1 classid 1:11 htb  prio 1 rate 1843Kbit ceil 122                                                                            88Kbit burst 3200b cburst 14400b quantum 23037
class change dev br0 parent 1:1 classid 1:12 htb  prio 2 rate 15360Kbit ceil 512                                                                            00Kbit burst 6400b cburst 64000b quantum 192000
class change dev eth0 parent 1:1 classid 1:12 htb  prio 2 rate 1228Kbit ceil 122                                                                            88Kbit burst 3200b cburst 14400b quantum 15350
class change dev br0 parent 1:1 classid 1:13 htb  prio 3 rate 10240Kbit ceil 512                                                                            00Kbit burst 6400b cburst 64000b quantum 128000
class change dev eth0 parent 1:1 classid 1:13 htb  prio 3 rate 2457Kbit ceil 122                                                                            88Kbit burst 3200b cburst 14400b quantum 30712
class change dev br0 parent 1:1 classid 1:14 htb  prio 4 rate 2560Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 32000
class change dev eth0 parent 1:1 classid 1:14 htb  prio 4 rate 614Kbit ceil 1228                                                                            8Kbit burst 3200b cburst 14400b quantum 7675
class change dev br0 parent 1:1 classid 1:15 htb  prio 5 rate 5120Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 64000
class change dev eth0 parent 1:1 classid 1:15 htb  prio 5 rate 1228Kbit ceil 122                                                                            88Kbit burst 3200b cburst 14400b quantum 15350
class change dev br0 parent 1:1 classid 1:16 htb  prio 6 rate 2560Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 32000
class change dev eth0 parent 1:1 classid 1:16 htb  prio 6 rate 614Kbit ceil 1228                                                                            8Kbit burst 3200b cburst 14400b quantum 7675
class change dev br0 parent 1:1 classid 1:17 htb  prio 7 rate 5120Kbit ceil 5120                                                                            0Kbit burst 6400b cburst 64000b quantum 64000
class change dev eth0 parent 1:1 classid 1:17 htb  prio 7 rate 3686Kbit ceil 122                                                                            88Kbit burst 3200b cburst 14400b quantum 46075

 

[dave14305]

Enable fq_codel in the Customize section of the FlexQoS GUI and test again for comparison.

 

[hancox]

I've always been in the same boat on upload only - glad to see it was never just me!

v1.2.1 with fq_codel has made this the best it's ever been, though - i think @dave14305 may have squashed this.

 

[mikefrommanchester]

Enable fq_codel in the Customize section of the FlexQoS GUI and test again for comparison.

Thank you Dave - that is significantly better at around 24ms on average. It still seems high but it's not terrible. Is there any other further optimisations I can do?

 

[mikefrommanchester]

After some more testing - the Upload BufferBloat is hugely variable as you can see from this GIF.... over 300ms.

I lowered my Upload to 5Mbps... but no difference.

My line stats are as follows:

 

[hancox]

After some more testing - the Upload BufferBloat is hugely variable as you can see from this GIF.... over 300ms.

I lowered my Upload to 5Mbps... but no difference.

My line stats are as follows:

View attachment 31029

This matches my experience exactly - it bounces like mad.

 

[mikefrommanchester]

This matches my experience exactly - it bounces like mad.

So essentially Asus QoS doesn't work on this router? Even with third party scripts that have been written by people cleverer than me, to help resolve this issue.

I guess I'll be sending it back then because the most important thing for me is to have low ping, stable connection for gaming.

It's strange, my connection is better with an old BT homehub flashed with OpenWRT...

Oh well, it's a bit annoying - I thought this RTAC86U would be a good upgrade but it seems it's actually a downgrade...

 

[bluzfanmr1]

To me, dslreports.com is very inconsistent and flaky. Have you tried another site? What about fast.com? See how close your loaded and unloaded latency are.

 

[BreakingDad]

dslreports can give a B C B rating, then run it again and it can give an A A A. Not a reliable or consistent site.

 

[underdose]

For the consistence of your DSLReports results, you might want to adjust your test settings as follows:

Also, with those relatively low speeds try switching to CakeQoS. It works far better than AdvancedQoS/FlexQoS on my RT-AX58U with 100DL/5UL connection and it is available for your router too!

GitHub - ttgapers/cakeqos-merlin: Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations.

Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations. - ttgapers/cakeqos-merlin

github.com

 

[Wade Coxon]

Agreed. Cake works a treat on the AC86U! A+ all round on a sub-100Mbps connection.

 

[mikefrommanchester]

For the consistence of your DSLReports results, you might want to adjust your test settings as follows:

View attachment 31076

Also, with those relatively low speeds try switching to CakeQoS. It works far better than AdvancedQoS/FlexQoS on my RT-AX58U with 100DL/5UL connection and it is available for your router too!

GitHub - ttgapers/cakeqos-merlin: Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations.

Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations. - ttgapers/cakeqos-merlin

github.com

Thanks - I will try this.

Fast.com were the same as DSLreports... I never had a problem with DSLReports in the past but I appreciate it can be down to the service provider.

 

[dave14305]

On my previous router with LuCi I had configured Cake successfully and BufferBloat was far better.

You will find far more expertise at OpenWRT regarding SQM and CAKE than you will here. And in general, you will find much newer kernels available under OpenWRT than ASUS and Broadcom make use of. So we make the best of the constraints we find ourselves under.

If CAKE works better for you, that will be great to hear what settings you decided to use.

 

[Morris]

If you are concerned about bufferbloat, use Cake QOS.

Morris

 

[mikefrommanchester]

+For the consistence of your DSLReports results, you might want to adjust your test settings as follows:

View attachment 31076

Also, with those relatively low speeds try switching to CakeQoS. It works far better than AdvancedQoS/FlexQoS on my RT-AX58U with 100DL/5UL connection and it is available for your router too!

GitHub - ttgapers/cakeqos-merlin: Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations.

Custom add-on for supported Asus routers with Merlin firmware that introduces CAKE QoS as an alternative to Traditional/Adaptive/FreshJR implementations. - ttgapers/cakeqos-merlin

github.com

Cake made a huge difference:

A+ A+ A+

Thank you!

 

[mikefrommanchester]

Also I notice in the Asus Adaptive QoS settings that the QoS whilst using Cake can remain disabled and seems to be forced into this mode. Is this normal?

If so, what a bonus. I'm really not sure I'd like to give Trend Micro my internet data for 'Analysis'... and really didn't want to but thought there was no option.