ci-e2e: Use kernel 6.1 instead of 6.0

[brb]

The former is the long term stable - https://kernel.org

[brb]

/ci-e2e

[brb]

/ci-e2e

[brb]

Unfortunately legit failure after changing kernel from 6.0 to 6.1 on the '6' config:

2023-11-23T08:51:43.413001842Z level=fatal msg="Failed to compile XDP program" error="attaching XDP program to interface eth0: loading eBPF collection into the kernel: program tail_nodeport_nat_ingress_ipv6: load program: argument list too long: BPF program is too large. Processed 1000001 insn (863 line(s) omitted)" subsys=datapath-loader

         - name: '6'
            # renovate: datasource=docker depName=quay.io/lvh-images/kind
            kernel: '6.1-20231026.065108'
            kube-proxy: 'none'
            kpr: 'true'
            tunnel: 'vxlan'
            lb-mode: 'snat'
            egress-gateway: 'true'
            host-fw: 'true'
            lb-acceleration: 'testing-only'
            ingress-controller: 'true'

Disabling host-fw resolves the issue.

[julianwiedmann]

@brb want to give this another try, now that #29377 is in?

Stable backports, all the way back? ugh ...

[brb]

/ci-e2e

[brb]

/ci-e2e-upgrade

[brb]

@julianwiedmann Cool, resolved!

Stable backports, all the way back? ugh ...

Any concerns about it?

[julianwiedmann]

Stable backports, all the way back? ugh ...

Any concerns about it?

Let's see which ones will need #29377, no point in worrying about it in advance 🤞 .

[julianwiedmann]

updating the labels, as

• The workflows in v1.12 have no reference to 6.0 kernels.
• The parts in v1.13 have been addressed by chore(deps): update stable lvh-images (v1.13) (patch) #30812.
• The parts in v1.14 were updated by chore(deps): update stable lvh-images (v1.14) (patch) #30653. But looks like we applied a s/6.0/6.6 instead of s/6.0/6.1. @lmb got a moment to restore harmony across the different branches? 🙏