-
Bug
-
Resolution: Done-Errata
-
Major
-
None
-
4.17
-
Quality / Stability / Reliability
-
False
-
None
-
3
-
Critical
-
No
-
None
-
Proposed
-
uShift Sprint 259
-
1
-
None
-
Release Note Not Required
-
N/A
-
None
-
None
-
None
-
None
Description of problem:
Setup microshift ipv6 singlestack cluster, created loadbalancer service with ETP and ITP 'cluster' on it. Sent traffic from buildvm to the external-ip, traffic was dropped. Meanwhile, sent traffic to endpoints passed, and traffic from testpod inside the cluster also passed.
Version-Release number of selected component (if applicable):
4.17
How reproducible:
always
Steps to Reproduce:
1. setup ipv6 singlestack microshift cluster
2. create loadbalancer svc on it.
$ oc get svc -o yaml
apiVersion: v1
items:
- apiVersion: v1
kind: Service
metadata:
creationTimestamp: "2024-08-30T10:44:08Z"
labels:
name: test-service
name: lbtest-etp-itp-cluster
namespace: e2e-ushift-sdn-60968-nnnrjaz9
resourceVersion: "103616"
uid: 3d0fed90-0076-4fc7-b200-600f23f56dba
spec:
allocateLoadBalancerNodePorts: true
clusterIP: fd02::a2e9
clusterIPs:
- fd02::a2e9
externalTrafficPolicy: Cluster
internalTrafficPolicy: Cluster
ipFamilies:
- IPv6
ipFamilyPolicy: SingleStack
ports:
- name: http
nodePort: 31721
port: 27017
protocol: TCP
targetPort: 8080
selector:
name: hello-pod
sessionAffinity: None
type: LoadBalancer
status:
loadBalancer:
ingress:
- ip: 2620:52:0:1eb:a117:863e:a7d6:82cc
ipMode: VIP
kind: List
metadata:
resourceVersion: ""
$ oc get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
hello-pod-0 1/1 Running 0 30m fd01:0:0:1::1c microshift-dev.local <none> <none>
hello-pod-1 1/1 Running 0 29m fd01:0:0:1::1d microshift-dev.local <none> <none>
test-pod 1/1 Running 0 29m fd01:0:0:1::1e microshift-dev.local <none> <none>
$ oc get svc
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
lbtest-etp-itp-cluster LoadBalancer fd02::cfb7 2620:52:0:1eb:a117:863e:a7d6:82cc 27017:30879/TCP 24s
3. curl from buildvm (the host where the microshift installed) or node failed. Meanwhile, curl from test-pod to external ip passed.
curl from buildvm to endpoints also passed.
$ timeout 5 curl -k [2620:52:0:1eb:a117:863e:a7d6:82cc]:27017
$ oc rsh test-pod curl -k [2620:52:0:1eb:a117:863e:a7d6:82cc]:27017
Hello OpenShift!
$ curl -k [fd01:0:0:1::1c]:8080
Hello OpenShift!
[redhat@microshift-dev ~]$ curl -k [fd01:0:0:1::1d]:8080
Hello OpenShift!
Actual results:
traffic from external to loadbalancer svc external-ip failed.
Expected results:
traffic from external to loadbalancer svc external-ip should pass.
Additional info:
- blocks
-
OCPBUGS-41623 traffic from external to microshift ipv6 loadbalancer svc failed
-
- Closed
-
- is cloned by
-
OCPBUGS-41623 traffic from external to microshift ipv6 loadbalancer svc failed
-
- Closed
-
- links to
-
RHEA-2024:6124
Red Hat build of MicroShift 4.18.z bug fix and enhancement update