Health-probe failing on cilium connectivity test --multi-cluster

[Rammurthy5]

Hi @littlejo, this is very useful cookbook. thanks for creating this.

this is not an ISSUE, rather a call for help!! I had to create an issue as there was no option for Q&A. apologies 🙏🏻

I followed all these steps (except for the cluster launch) and when i run connectivity test between cluster1 and cluster2, i see health-probe fails.

I am using EKS v1.25 kubernetes, VPC peered, and security groups appropriately configured. I followed your guide for cilium installation -- https://github.com/littlejo/cilium-eks-cookbook/blob/main/install-cilium-eks-helm.md.
i jumped into cilium agents in both clusters, and ran cilium-health status, cilium-health status --probe -o=json

i am not sure where does 10.0.2.13 IP live, when i checked kubectl get pods -A -o wide, i dont see this IP at all in both clusters. do you know what am i missing out, please?

cluster2/ip-10-0-2-109.compute.internal:
  Host connectivity to 10.0.2.109:
   ICMP to stack:  OK, RTT=12.260828ms
   HTTP to agent:  OK, RTT=12.406438ms
  Endpoint connectivity to 10.0.2.13:
   ICMP to stack:  Connection timed out
   HTTP to agent:  Get "http://10.0.2.13:4240/hello": context deadline exceeded (Client.Timeout exceeded while awaiting headers)

if i run # cilium-health status --probe -o=json
Error: Cannot get status/probe: Put "http://%2Fvar%2Frun%2Fcilium%2Fhealth.sock/v1beta/status/probe": context deadline exceeded

EDIT:
cluster2 has all 4 endpoints (2 from cluster1, 2 from cluster2) reachable but cluser1 has 3 endpoints reachable and one unreachable (1 endpoint from cluster1 is reachable but other endpoint in cluster1 is unreachable)

[littlejo]

Hello,

i'm not sure to understand what you want to do. You want create a clustermesh between 2 clusters? If yes, the guide is https://github.com/littlejo/cilium-eks-cookbook/blob/main/install-cilium-eks-clustermesh.md and not https://github.com/littlejo/cilium-eks-cookbook/blob/main/install-cilium-eks-helm.md.

[Rammurthy5]

@littlejo , thanks for getting back. yes, thats right i followed the same must be typo in the previous message.
Unfortunately, not all endpoints are reachable for one of my clusters. I have tried it on kubernetes 1.25, 1.26, 1.27, 1.28.2 but nothing helps. may i ask if you could help, please?