[v2.0.0 Issue]: panic: runtime error: invalid memory address or nil pointer dereference

[CoreTechnologyAG]

What happened?
Kube-Router does not start after starting with v2.0.0 and afterthat the whole kubernetes cluster aren't working.

If we start the kube-router with the latest v1.6 version, it works fine.

What did you expect to happen?
Starting Kube-Router V2.0 as it did in V1.6

How can we reproduce the behavior you experienced?
Steps to reproduce the behavior:

1. Start Kube-router based on https://raw.githubusercontent.com/cloudnativelabs/kube-router/master/daemonset/kubeadm-kuberouter-all-features.yaml

Screenshots / Architecture Diagrams / Network Topologies

• kubernetes cluster with
  • 3 Base-Server with etcd and coredns
  • 2 Kubernetes Master-Server kubeadm, kubelet, kubectl: GitVersion:"v1.26.6",
  • 3 Kubernetes Node-Server

System Information (please complete the following information):

• Kube-Router Version: v2.0.0 (Latest)
• Kube-Router Parameters:
  - --run-router=true
  - --run-firewall=true
  - --run-service-proxy=true
  - --bgp-graceful-restart=true
  - --kubeconfig=/var/lib/kube-router/kubeconfig
• Kubernetes Version: 1.26.6-00 on Ubuntu 20.04.6 LTS
• Cloud Type: on premise
• Kubernetes Deployment Type: Kubeadm
• Kube-Router Deployment Type: DaemonSet
• Cluster Size: 3 Nodes

Logs, other output, metrics
I1012 08:14:55.591795 1 network_policy_controller.go:163] Starting network policy controller
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x98 pc=0x188395e]
goroutine 154 [running]:
github.com/cloudnativelabs/kube-router/v2/pkg/controllers/proxy.(*NetworkServicesController).deleteHairpinIptablesRules(0xc00070b878?, {0x22d179c?, 0x3d?})
/build/pkg/controllers/proxy/network_services_controller.go:1361 +0x5e
github.com/cloudnativelabs/kube-router/v2/pkg/controllers/proxy.(*NetworkServicesController).syncHairpinIptablesRules(0xc000512c40)
/build/pkg/controllers/proxy/network_services_controller.go:1279 +0x33e
github.com/cloudnativelabs/kube-router/v2/pkg/controllers/proxy.(*NetworkServicesController).doSync(0xc000512c40)
/build/pkg/controllers/proxy/network_services_controller.go:374 +0x210
github.com/cloudnativelabs/kube-router/v2/pkg/controllers/proxy.(*NetworkServicesController).Run(0xc000512c40, 0x0?, 0xc0000a0360, 0xc00005ef40)
/build/pkg/controllers/proxy/network_services_controller.go:287 +0x57a
created by github.com/cloudnativelabs/kube-router/v2/pkg/cmd.(*KubeRouter).Run
/build/pkg/cmd/kube-router.go:196 +0x10aa

Thanks for your work at all for the amazing job.
Ritzo

[aauren]

Have you read and followed the release instructions for v2.0.0? https://github.com/cloudnativelabs/kube-router/releases/tag/v2.0.0

[flababah]

FYI, I ran into the same problem today trying to create a new cluster from scratch following
https://github.com/cloudnativelabs/kube-router/blob/master/docs/kubeadm.md

The yaml file seems to contain the new ClusterRole RBAC updates for endpointslices (da73dea) that were mentioned in the release instructions.

The setup is as basic as it can get. Single node, kubeadm init --pod-network-cidr='192.168.0.0/16' and kubeadm-kuberouter-all-features.yaml applied. Same nil pointer dereference. Any ideas?

kubeadm-kuberouter.yaml (--run-service-proxy=false) works.

[aauren]

I think that I've found the issue. Working on a fix now.