Optimizing smp_affinity for irqs on interfaces under conditions of max load

How does one know which CPU/core cake is bound to or even, how would one move it around? So far I was just trying to distribute irqs around so that no CPU ended up hitting 100% under bw saturation.

The default WAN port handle "duplex traffic". The only way this would be beneficial is to split the Ingress & Egress directions OpenWRT main router w/o WiFi.

Example:

WAN ----→ eth0------→ Egress [ Cake Upload ]

LAN -----→ eth1-----→ Ingress [ Cake Download ] (No IFB)

Hello everyone, at the moment I have an archer c6 v3 with the 23.05 update, as it is a router with 2 cores that are divided into 2 more, I'm here from Brazil and I'm translating this post, I'm at war with my router because I can't activate 2 cores to work on the eth line, whenever I activate a core or CPU0 for CPU1, CPU0 is null and the work is on CPU1, I will leave the commands I made here that I got from this site: https://openwrt.org/docs/guide-user/advanced/load_balancing_-_tuning_smp_irq

#eth0 core 0
echo 8 > /proc/irq/21/smp_affinity

#queues on all cores

echo f > /sys/class/net/eth1/queues/rx-0/rps_cpus

because I believe there are settings that I don't know about, as I'm new to this journey.