kube-static-egress-ip
kube-static-egress-ip copied to clipboard
Failed to setup routes
logs
I0723 19:18:46.507067 1 director.go:102] Created ipset name: EGRESS-IP-XFVL3XZHQWBKAPWE
I0723 19:18:46.507676 1 director.go:111] Added ips [192.168.7.10] to the ipset name: EGRESS-IP-XFVL3XZHQWBKAPWE
I0723 19:18:46.508735 1 director.go:127] iptables rule in mangle table PREROUTING chain to match src to ipset
E0723 19:18:46.522160 1 controller.go:286] Failed to setup routes to send the egress traffic to gateway due to Failed to add route in custom route table due to: exit status 2
I0723 19:18:46.522179 1 controller.go:199] Successfully synced 'resequip/test'
manifest
apiVersion: staticegressips.nirmata.io/v1alpha1
kind: StaticEgressIP
metadata:
name: test
spec:
rules:
- egressip: 10.35.12.65
service-name: backend
cidr: 10.35.12.17/32
service
apiVersion: v1
kind: Service
metadata:
annotations:
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"labels":{"app":"backend"},"name":"backend","namespace":"resequip"},"spec":{"ports":[{"name":"http","port":80}],"selector":{"app":"backend"}}}
metallb.universe.tf/address-pool: oebs
creationTimestamp: "2019-07-12T01:43:53Z"
labels:
app: backend
name: backend
namespace: resequip
resourceVersion: "249156023"
selfLink: /api/v1/namespaces/resequip/services/backend
uid: 81891851-a446-11e9-96c8-0050562c0156
spec:
clusterIP: 192.168.241.54
externalTrafficPolicy: Local
healthCheckNodePort: 32592
loadBalancerIP: 10.35.12.65
ports:
- name: http
nodePort: 30342
port: 80
protocol: TCP
targetPort: 80
selector:
app: backend
sessionAffinity: None
type: LoadBalancer
status:
loadBalancer:
ingress:
- ip: 10.35.12.65
I have the same problem, someone help?
same problem here, any news about this?
I think i debugged it to this line: https://github.com/nirmata/kube-static-egress-ip/blob/master/pkg/director/director.go#L148
Connected to container, tried to add routes manually, but ip route
yields RTNETLINK answers: Network is unreachable.
even though i am able to ping the host (same network) and even connect via ssh.
I start to think it's some GKE specific thing (i am using GKE for this), probably the node subnet address does not allow setting the other node as gateway for the route.