QoSmate: (Yet Another) Quality of Service Tool for OpenWrt

Hudra · 2024-12-25T10:51:32.105Z

Knomax:

Hudra:

No, please try:

tcpdump -i eth0.2 -v -n icmp

Nothing to write.

Okay, please try the following command:

tcpdump -i pppoe-wan -v -n -Q in icmp

I did some more research and gave the problem additional thought. Here are a few ideas and observations:

Fundamentally, it's hard for me to reproduce the issue because I have multiple test devices running different OpenWRT versions, and I don't experience the problem on any of them.

It seems like your device might still be using swconfig instead of DSA. I’m not sure if that could have any impact, but it’s worth considering.

Please double-check if you’ve truly disabled both software and hardware offloading.

At the beginning, I suggested using the postrouting nftables hook. I think you said you have tried that? Here are some additional ideas related to the nftables hook and priority:

Try experimenting with different combinations (you can configure this under the advanced section of the UI):

forward + -150
postrouting + 0
postrouting + -150

After each change, verify if the DSCP values are being written into the conntrack table.

Although you’ve already tried the 24.10 release, you could also consider testing a master snapshot build. These often include newer nftables versions, which might resolve your issue.

Another option is to consider upgrading to a newer device.

Finally, could you share the output of the following commands with us?

nft -c -d netlink add rule inet dscptag dscptag ct mark set ip dscp or 128 counter
nft -c -d netlink add rule inet dscptag dscptag ct mark set ip6 dscp or 128 counter

nft -c -d netlink add rule inet dscptag dscptag meta nfproto ipv4 ct mark set ip dscp or 128 counter
nft -c -d netlink add rule inet dscptag dscptag meta nfproto ipv6 ct mark set ip6 dscp or 128 counter

Maybe @brada4 has some additional ideas?

brada4 · 2024-12-25T14:26:29.864Z

It is rather strange to get such poor results. If you generate significant traffic on the router qosmate should go to postrouting. In either case it is best to have it AFTER default tables that could drop/reject connections to get less data for us.

Another isolated test would be to nft fush table inet dscpmark to remove qosmate firewall part, just to figure our which half is at fault.
i.e if latency normalizes if firewall part is removed firewall rules are at fault
if nothing changes then it is qdisc chaining.

(also admit if you have more than dozen rules/checkboxes added/changed to default fw4 one is usually not kill

the nftables part adds 30 bytes to 200 bytes of tracked state memory use and about 10 rules to 50 in fw4. In summary it is definitely slower, but not by tragic degree.
I have tested mostly of 100mbps devices which now does not exceed SoC CPU/RAM of the time. So 100/100 (102400kbps) should work fine with some breathing space left on CPU, scaling linearily.

Hudra · 2024-12-25T15:42:15.523Z

brada4:

to get such poor results

brada4:

i.e if latency normalizes

I believe the issue @Knomax is experiencing (which I referred to in my previous post) is not related to performance or latency but rather that the DSCP value is simply not being written to the conntrack. If you take another look at the discussion, you'll see that his egress traffic is correctly marked, but the value is not being written to the conntrack, and that’s what my last post was addressing.

antonk · 2024-12-25T16:38:55.791Z

Knomax:

                meta nfproto 0 ct mark set @nh,8,8 & 0xfc [invalid type] | 0x80 counter packets 0 bytes 0
                meta nfproto 0 ct mark set @nh,0,16 & 0xfc0 [invalid type] | 0x80 counter packets 0 bytes 0

I suppose [invalid type] is not expected here? @Hudra

brada4 · 2024-12-25T16:40:46.034Z

Looks quite sane?

inet dscptag dscptag
  [ meta load nfproto => reg 1 ]
  [ cmp eq reg 1 0x0000000a ]
  [ payload load 2b @ network header + 0 => reg 1 ]
  [ bitwise reg 1 = ( reg 1 & 0x0000c00f ) ^ 0x00000000 ]
  [ byteorder reg 1 = ntoh(reg 1, 2, 2) ]
  [ bitwise reg 1 = ( reg 1 >> 0x00000006 ) ]
  [ bitwise reg 1 = ( reg 1 & 0xffffff7f ) ^ 0x00000080 ] <--- endiannes-swapped
  [ ct set mark with reg 1 ] <--- no mistake here....
  [ counter pkts 0 bytes 0 ]

What I was alluding to (in output of nft -c -d netlink -f dscptag.nft)

  [ payload load 2b @ network header + 0 => reg 1 ]
  [ bitwise reg 1 = ( reg 1 & 0x000003ff ) ^ 0x00008000 ]
  [ payload write reg 1 => 2b @ network header + 0 csum_type 1 csum_off 10 csum_flags 0x0 ]

ie that checksum is recalculated on invalid packets then null state leaks into our dynamic tables and something chaotic is restored on ingress... (and the performance effect of calculating checksum not updating via bitops ?)

Hudra · 2024-12-25T18:10:46.765Z

Good catch, but this is expected behavior in nftables versions < v1.0.9 and it should still work. Starting from version v1.0.9, [invalid type] will no longer be displayed.

Read here:

CAKE w/ DSCPs - cake-qos-simple Community Builds, Projects & Packages

I just updated to this version and noticed in the output of nft list ruleset chain store-dscp-in-conntrack { meta nfproto ipv4 ct mark set @nh,8,8 & 0xfc [invalid type] I found mention of it in this bug report[1]. I’m not sure of its ramifications. Is this an issue? [1] https://www.spinics.net/lists/netfilter/msg61171.html

Hudra · 2024-12-25T18:52:15.684Z

brada4:

Looks quite sane?

Yes, my output looks exactly the same as yours, and everything works for me as well.

brada4:

  [ bitwise reg 1 = ( reg 1 & 0x0000c00f ) ^ 0x00000000 ]
  [ byteorder reg 1 = ntoh(reg 1, 2, 2) ]
  [ bitwise reg 1 = ( reg 1 >> 0x00000006 ) ]
  [ bitwise reg 1 = ( reg 1 & 0xffffff7f ) ^ 0x00000080 ] <--- endiannes-swapped

@brada4 do you think it could be an endian issue? We also suspected this in the past when I troubleshooted with @Ridz :

QOSMate / Cake w/ DSCPs Network and Wireless Configuration

root@OpenWrt:~# nft -c -d netlink add rule inet dscptag dscptag ct mark set ip dscp or 128 counter inet dscptag dscptag [ meta load nfproto => reg 1 ] [ cmp eq reg 1 0x02000000 ] [ payload load 1b @ network header + 1 => reg 1 ] [ bitwise reg 1 = ( reg 1 & 0xfc000000 ) ^ 0x00000000 ] [ bitwise reg 1 = ( reg 1 >> 0x00000002 ) ] [ bitwise reg 1 = ( reg 1 & 0xffffff7f ) ^ 0x00000080 ] [ ct set mark with reg 1 ] [ counter pkts 0 bytes 0 ] root@OpenWrt:~# nft -c -d netlink add rule i…

@Knomax, could you please provide the following additional output along with the instructions I gave you above?

##check endian
[ "$(printf '\1\0\0\0' | hexdump -v -e '/1 "%02x"')" = "01000000" ] && echo "little" || echo "big"

brada4 · 2024-12-25T19:56:56.926Z

Not endianness issue, just curiously long ritual with bitops.

oneinchashh · 2024-12-26T03:24:14.170Z

Thank you for this wonderful package. May I ask if it can possibly handle 2 isp wans by different configs for two wan interfaces? I'm using mwan3 btw.

Hudra · 2024-12-26T09:30:56.330Z

No, unfortunately, QoSmate currently only works with a single WAN interface. Do you primarily use mwan3 for load balancing or to route specific traffic through a particular WAN? If you can provide more details, maybe we can find a workaround.

oneinchashh · 2024-12-26T09:49:26.916Z

for active/backup only, my two isps have different speed

brada4 · 2024-12-26T10:42:32.989Z

You have to add extra "egress wash" rule - the oifname "wan" jump mark_cs0 you see at the end of nft list ruleset, and set some priority-aware qdisc on new interface pfifo_fast < hfsc < cake
Thats not "full support" , but at least make sure qosmate does not do extra damage.

wolvth · 2024-12-27T02:33:52.970Z

My system is running ImmortalWRT 23.5.03 on a NanoPi R4S. uname -a 5.15.162
QOSmate fails to autostart upon system reboot and requires manual activation.There are no other issues.

QOSmate version Backend versions:
Current version: 0.5.42
Frontend versions:
Current version: 1.0.12

Yesterday, I upgraded the Frontend versions to the latest version 1.0.13. After that, QOSmate can autostart,
but the machine suddenly reboots automatically. Everything is set to default, and this has happened 3 times.
I've now set the Root Queueing Discipline to cake and will continue monitoring
It seems like the reboots happen shortly after speed tests.I managed to get some logs.

Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.251728]   ESR = 0x0000000096000006
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.252075]   EC = 0x25: DABT (current EL), IL = 32 bits
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.252548]   SET = 0, FnV = 0
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.252823]   EA = 0, S1PTW = 0
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.253107]   FSC = 0x06: level 2 translation fault
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.253799]   ISV = 0, ISS = 0x00000006
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.254147]   CM = 0, WnR = 0
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.254415] swapper pgtable: 4k pages, 48-bit VAs, pgdp=0000000004272000
Thu Dec 26 10:35:51 2024 kern.alert kernel: [  593.255008] [ffff0000000005b2] pgd=180000003fff8003, p4d=180000003fff8003, pud=180000003fff7003, pmd=0000000000000000
Thu Dec 26 10:35:51 2024 kern.emerg kernel: [  593.255953] Internal error: Oops: 0000000096000006 [#1] PREEMPT SMP
Thu Dec 26 10:36:22 2024 kern.warn kernel: [   11.544116] rk_gmac-dwmac fe300000.ethernet eth0: PTP not supported by HW
Thu Dec 26 10:36:23 2024 daemon.warn avahi-daemon[2039]: WARNING: No NSS support for mDNS detected, consider installing nss-mdns!
Thu Dec 26 10:36:23 2024 daemon.err collectd[2186]: Initialization complete, entering read-loop.
Thu Dec 26 10:36:30 2024 daemon.err nmbd[3184]: [2024/12/26 10:36:30.073991,  0] ../../source3/nmbd/nmbd.c:901(main)
Thu Dec 26 10:36:30 2024 daemon.err nmbd[3184]:   nmbd version 4.18.8 started.
Thu Dec 26 10:36:30 2024 daemon.err nmbd[3184]:   Copyright Andrew Tridgell and the Samba Team 1992-2023
Thu Dec 26 10:36:30 2024 daemon.err smbd[3183]: [2024/12/26 10:36:30.119012,  0] ../../source3/smbd/server.c:1746(main)
Thu Dec 26 10:36:30 2024 daemon.err smbd[3183]:   smbd version 4.18.8 started.
Thu Dec 26 10:36:30 2024 daemon.err smbd[3183]:   Copyright Andrew Tridgell and the Samba Team 1992-2023
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.096394] Unable to handle kernel paging request at virtual address ffff0000000005b2
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.097362]   ESR = 0x0000000096000006
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.097699]   EC = 0x25: DABT (current EL), IL = 32 bits
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.098172]   SET = 0, FnV = 0
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.098448]   EA = 0, S1PTW = 0
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.098730]   FSC = 0x06: level 2 translation fault
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.099423]   ISV = 0, ISS = 0x00000006
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.099767]   CM = 0, WnR = 0
Thu Dec 26 10:37:26 2024 kern.alert kernel: [   75.100035] swapper pgtable: 4k pages, 48-bit VAs, pgdp=0000000004272000
Thu Dec 26 10:37:57 2024 kern.warn kernel: [   11.533278] rk_gmac-dwmac fe300000.ethernet eth0: PTP not supported by HW
Thu Dec 26 10:37:58 2024 daemon.warn avahi-daemon[2049]: WARNING: No NSS support for mDNS detected, consider installing nss-mdns!
Thu Dec 26 10:37:58 2024 daemon.err collectd[2173]: Initialization complete, entering read-loop.
Thu Dec 26 10:38:05 2024 daemon.err nmbd[3183]: [2024/12/26 10:38:05.724406,  0] ../../source3/nmbd/nmbd.c:901(main)
Thu Dec 26 10:38:05 2024 daemon.err nmbd[3183]:   nmbd version 4.18.8 started.
Thu Dec 26 10:38:05 2024 daemon.err nmbd[3183]:   Copyright Andrew Tridgell and the Samba Team 1992-2023
Thu Dec 26 10:38:05 2024 daemon.err smbd[3182]: [2024/12/26 10:38:05.765372,  0] ../../source3/smbd/server.c:1746(main)
Thu Dec 26 10:38:05 2024 daemon.err smbd[3182]:   smbd version 4.18.8 started.
Thu Dec 26 10:38:05 2024 daemon.err smbd[3182]:   Copyright Andrew Tridgell and the Samba Team 1992-2023
Thu Dec 26 10:38:06 2024 cron.err crond[5799]: crond (busybox 1.36.1) started, log level 5
Thu Dec 26 10:38:07 2024 daemon.warn odhcpd[1569]: No default route present, setting ra_lifetime to zero!
Thu Dec 26 10:38:11 2024 daemon.warn odhcpd[1569]: No default route present, setting ra_lifetime to zero!
Thu Dec 26 10:38:11 2024 daemon.warn pppd[3720]: Connected to 00:18:82:d0:21:f6 via interface eth0

brada4 · 2024-12-27T06:09:40.692Z

Maybe you format </> your log if you want ayone to read it?

brada4 · 2024-12-27T07:09:13.962Z

It has to be started once via command line, then after reboot run logread -e qosmate

Hudra · 2024-12-27T08:55:59.972Z

Officially, only OpenWrt is supported. Forks like ImmortalWRT may have fundamental changes or adjustments that are difficult for me to track. Nevertheless, we’ve encountered similar issues in the past (primarily with users using ImmortalWRT). I made adjustments to the init script back then, which seemed to resolve the problem for those users.

For context: qosmate uses a hotplug script that triggers on the UP status of network devices. This means that if an interface goes down and is restarted by the system, the qosmate hotplug script will kick in and restart qosmate. This can happen multiple times during startup, especially when using PPPoE/PPPoA. So, it’s possible that there’s no real issue, and you may have simply checked the status of qosmate while it was restarting or before it had fully started.

I recommend trying a few reboots. Once OpenWrt has fully booted, wait 2-3 minutes before checking whether qosmate is active or not.

As for frontend updates, they should have no impact on qosmate’s autostart. The changes from version 1.0.12 to 1.0.13 were limited to enhancements to the filtering logic in the connections tab.

Things to Check:

You might want to see if there’s a race condition with other applications during startup.

Regarding the other issue of qosmate rebooting itself:

The logs you posted look more like a kernel-level memory fault, not a user-space application crash. This leads me to believe that qosmate isn’t directly causing the issue.

A few thoughts:

If possible, consider switching to the official OpenWrt version.
Update ImmortalWRT:
It looks like you’re not even running the latest version of ImmortalWRT. I’d recommend updating to the latest stable version and testing again. If the issue persists, you could also try updating to version 24.10.
Reboots during speed tests:
You mentioned the reboots occur shortly after running speed tests. This likely indicates that you’re using the auto-setup function. Note that the auto-setup function only needs to be used once, not repeatedly. Its purpose is to simplify the initial configuration and should ideally only be used one time.The auto-setup function uses either speedtest-go (if sufficient storage is available) or python3-speedtest-cli. It’s possible that the installed package is causing the issue. You can test this by identifying which package is installed and manually running a speed test to see if the crash can be reproduced.Additionally, during auto-setup, a file is written to /tmp (your RAM). This might be causing the problem. You could test this by checking if writing to /tmp triggers a failure.

To check if you might have configured something incorrectly, could you please also share the following output:

/etc/init.d/qosmate status
cat /etc/config/qosmate

Knomax · 2024-12-27T10:24:15.678Z

Hi..how to change to DSA configuration?It isnt enabled by default in latest Openwrt versions?My connection is pppoe if it matters.

wolvth · 2024-12-27T10:36:41.178Z

Hudra:

/etc/init.d/qosmate status

root@ImmortalWrt:~# /etc/init.d/qosmate status
==== qosmate Status ====
qosmate service is enabled.
Traffic shaping (HFSC) is active on the egress interface (pppoe-wan).
Traffic shaping is active on the ingress interface (ifb-pppoe-wan).
==== Overall Status ====
qosmate is currently active and managing traffic shaping.
==== Current Settings ====
Upload rate: 60000 kbps
Download rate: 900000 kbps
Game traffic upload: 9400 kbps
Game traffic download: 135400 kbps
Queue discipline: pfifo (for game traffic in HFSC)
==== Package Status ====
All required packages are installed.

==== Detailed Technical Information ====
Traffic Control (tc) Queues:
qdisc noqueue 0: dev lo root refcnt 2
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc mq 0: dev eth0 root
 Sent 73463964 bytes 324243 pkt (dropped 0, overlimits 0 requeues 2)
 backlog 0b 0p requeues 2
qdisc fq_codel 0: dev eth0 parent :1 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 4Mb ecn drop_batch 64
 Sent 73463964 bytes 324243 pkt (dropped 0, overlimits 0 requeues 2)
 backlog 0b 0p requeues 2
  maxpacket 1486 drop_overlimit 0 new_flow_count 4 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc mq 0: dev eth1 root
 Sent 1322106682 bytes 1062002 pkt (dropped 0, overlimits 0 requeues 1965)
 backlog 0b 0p requeues 1965
qdisc fq_codel 0: dev eth1 parent :1 limit 10240p flows 1024 quantum 1514 target 5ms interval 100ms memory_limit 4Mb ecn drop_batch 64
 Sent 1322106682 bytes 1062002 pkt (dropped 0, overlimits 0 requeues 1965)
 backlog 0b 0p requeues 1965
  maxpacket 27252 drop_overlimit 0 new_flow_count 26781 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc noqueue 0: dev br-lan root refcnt 2
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc clsact ffff: dev br-lan parent ffff:fff1
 Sent 618004264 bytes 768888 pkt (dropped 484, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc noqueue 0: dev dae0 root refcnt 2
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc clsact ffff: dev dae0 parent ffff:fff1
 Sent 475600448 bytes 411671 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc hfsc 1: dev pppoe-wan root refcnt 2 default 13
 Sent 236703 bytes 962 pkt (dropped 0, overlimits 58 requeues 0)
 backlog 0b 0p requeues 0
qdisc fq_codel 8005: dev pppoe-wan parent 1:12 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 1500000b ecn drop_batch 64
 Sent 4598 bytes 49 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 137 drop_overlimit 0 new_flow_count 24 ecn_mark 0
  new_flows_len 1 old_flows_len 0
qdisc fq_codel 8007: dev pppoe-wan parent 1:14 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 1500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc pfifo 10: dev pppoe-wan parent 1:11 limit 405p
 Sent 96 bytes 1 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc fq_codel 8006: dev pppoe-wan parent 1:13 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 1500000b ecn drop_batch 64
 Sent 232009 bytes 912 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 2952 drop_overlimit 0 new_flow_count 422 ecn_mark 0
  new_flows_len 1 old_flows_len 6
qdisc fq_codel 8008: dev pppoe-wan parent 1:15 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 1500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc clsact ffff: dev pppoe-wan parent ffff:fff1
 Sent 636703410 bytes 822799 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc hfsc 1: dev ifb-pppoe-wan root refcnt 2 default 13
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc fq_codel 8009: dev ifb-pppoe-wan parent 1:12 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 22500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc fq_codel 800b: dev ifb-pppoe-wan parent 1:14 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 22500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc pfifo 10: dev ifb-pppoe-wan parent 1:11 limit 6005p
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
qdisc fq_codel 800a: dev ifb-pppoe-wan parent 1:13 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 22500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0
qdisc fq_codel 800c: dev ifb-pppoe-wan parent 1:15 limit 10240p flows 1024 quantum 3000 target 4ms interval 100ms memory_limit 22500000b ecn drop_batch 64
 Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0)
 backlog 0b 0p requeues 0
  maxpacket 0 drop_overlimit 0 new_flow_count 0 ecn_mark 0
  new_flows_len 0 old_flows_len 0

==== Nftables Ruleset (dscptag) ====
        chain dscptag {
                type filter hook forward priority filter; policy accept;
                counter packets 136 bytes 8088 jump mark_cs0
                meta l4proto udp ct original proto-src 51413 counter packets 0 bytes 0 jump mark_cs1
                meta l4proto udp ct original proto-dst 51413 counter packets 0 bytes 0 jump mark_cs1
                meta l4proto tcp ct original proto-dst { 6881-6889, 51413 } counter packets 0 bytes 0 jump mark_cs1
                meta length < 100 tcp flags ack add @xfst4ack { ct id . ct direction limit rate over 300000/second } counter packets 0 bytes 0 jump drop995
                meta length < 100 tcp flags ack add @fast4ack { ct id . ct direction limit rate over 30000/second } counter packets 0 bytes 0 jump drop95
                meta length < 100 tcp flags ack add @med4ack { ct id . ct direction limit rate over 3000/second } counter packets 0 bytes 0 jump drop50
                meta length < 100 tcp flags ack add @slow4ack { ct id . ct direction limit rate over 3000/second } counter packets 0 bytes 0 jump drop50
                meta l4proto tcp ct bytes < 56250000 jump mark_500ms
                meta l4proto tcp ct bytes > 1125000000 jump mark_10s
                meta l4proto tcp ip dscp != cs1 add @slowtcp { ct id . ct direction limit rate 150/second burst 150 packets } ip dscp set af42 counter packets 73 bytes 3746
                meta l4proto tcp ip6 dscp != cs1 add @slowtcp { ct id . ct direction limit rate 150/second burst 150 packets } ip6 dscp set af42 counter packets 54 bytes 3640
                meta l4proto { tcp, udp } th dport { 53, 853, 3389, 5353 } ip dscp set cs5 counter packets 0 bytes 0 comment "dns"
                ip saddr 192.168.2.133 udp dport != { 80, 443 } ip dscp set cs5 counter packets 1 bytes 56 comment "Game_Console_Outbound"
                udp sport != { 80, 443 } ip daddr 192.168.2.133 ip dscp set cs5 counter packets 0 bytes 0 comment "Game_Console_Inbound"
                meta priority set ip dscp map @priomap counter packets 82 bytes 4448
                meta priority set ip6 dscp map @priomap counter packets 54 bytes 3640
                meta nfproto ipv4 ct mark set @nh,8,8 & 0xfc [invalid type] | 0x80 counter packets 82 bytes 4448
                meta nfproto ipv6 ct mark set @nh,0,16 & 0xfc0 [invalid type] | 0x80 counter packets 54 bytes 3640
                oifname "pppoe-wan" jump mark_cs0
        }
}

==== Custom Rules Table Status ====
Custom rules table (qosmate_custom) is not active or doesn't exist.

root@ImmortalWrt:~# cat /etc/config/qosmate

config global 'global'
        option enabled '1'

config settings 'settings'
        option WAN 'pppoe-wan'
        option DOWNRATE '900000'
        option UPRATE '60000'
        option ROOT_QDISC 'hfsc'

config advanced 'advanced'
        option MSS '536'
        option PRESERVE_CONFIG_FILES '1'
        option WASHDSCPUP '1'
        option WASHDSCPDOWN '1'
        option BWMAXRATIO '20'
        option UDP_RATE_LIMIT_ENABLED '0'
        option TCP_UPGRADE_ENABLED '1'
        option UDPBULKPORT '51413'
        option TCPBULKPORT '51413,6881-6889'
        option NFT_HOOK 'forward'
        option NFT_PRIORITY '0'

config hfsc 'hfsc'
        option LINKTYPE 'ethernet'
        option OH '44'
        option gameqdisc 'pfifo'
        option nongameqdisc 'fq_codel'
        option nongameqdiscoptions 'besteffort ack-filter'
        option MAXDEL '24'
        option PFIFOMIN '5'
        option PACKETSIZE '450'
        option netemdelayms '30'
        option netemjitterms '7'
        option netemdist 'normal'
        option pktlossp 'none'

config cake 'cake'
        option COMMON_LINK_PRESETS 'ethernet'
        option PRIORITY_QUEUE_INGRESS 'diffserv4'
        option PRIORITY_QUEUE_EGRESS 'diffserv4'
        option HOST_ISOLATION '1'
        option NAT_INGRESS '1'
        option NAT_EGRESS '1'
        option ACK_FILTER_EGRESS 'auto'
        option AUTORATE_INGRESS '0'

config custom_rules 'custom_rules'

config rule
        option name 'dns'
        option class 'cs5'
        option counter '1'
        option enabled '1'
        option proto 'tcp udp'
        list dest_port '53'
        list dest_port '853'
        list dest_port '3389'
        list dest_port '5353'

config rule
        option name 'Game_Console_Outbound'
        option proto 'udp'
        option src_ip '192.168.2.133'
        list dest_port '!=80'
        list dest_port '!=443'
        option class 'cs5'
        option counter '1'
        option enabled '1'

config rule
        option name 'Game_Console_Inbound'
        option proto 'udp'
        option dest_ip '192.168.2.133'
        list src_port '!=80'
        list src_port '!=443'
        option class 'cs5'
        option counter '1'
        option enabled '1'

Hudra · 2024-12-27T10:40:28.105Z

Config looks good!

wolvth · 2024-12-27T10:42:36.738Z

I've switched back to HFSC and will keep monitoring.