Indeterministic behavior while switching gateways

[Tang8330]

As per title.

I have this set up such that any AWS or GCP resources will go through my egress gateway.

However, I would like to have some redundancies by having 2 gateways such that I can swap one out, reboot, etc...and achieve HA.

When I switch gateways, the mapping does not change and when running describe static route, it shows that none of the routes are ready.

Checking events is also extremely extremely slow / verging on not being able due to how many events there are.

For context, the number of CIDR ranges I am trying to load is ~2k (list from GCP and AWS). I have increased the memory limit for my staticRoute operator from 150Mi to 512Mi.

Any help would be appreciated.

[Tang8330]

One thing I also noticed -- when changing the gateway and then running describe kubectl describe staticroute NAME.

The K8 resource status still points to the old gateway.

[nabillarahmanizhafira]

Hi is there any update on this? 😓