-
Bug
-
Resolution: Unresolved
-
Blocker
-
None
-
None
-
None
-
Critical
-
Customer Escalated
-
2
-
rhel-sst-security-crypto
-
ssg_security
-
None
-
False
-
-
None
-
SDN Sprint 254, SDN Sprint 255
-
None
-
None
-
None
One control plane node ( here xiyun-20-m-zxpf6-control-plane-1) cannot access the pod on one worker node(xiyun-20-m-zxpf6-compute-0).The profile used in the clusters: /upi-on-vsphere/versioned-installer-ovn-ipsec-static_network-bonding-active-backup-secureboot-fips-hw19-ci
Version-Release number of selected component (if applicable):
4.14.0-0.nightly-2023-10-18-004928
How reproducible:
Sometimes
Steps to Reproduce:
1, create a pod on one worker node xiyun-20-m-zxpf6-compute-0 # oc apply -f pod.yaml pod/hello-pod created # oc get pod hello-pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES hello-pod 1/1 Running 0 40s 10.131.0.65 xiyun-20-m-zxpf6-compute-0 <none> <none> 2, Try to access this pod from master xiyun-20-m-zxpf6-control-plane-1, but failed to access it. # oc debug node/xiyun-20-m-zxpf6-control-plane-1 Temporary namespace openshift-debug-h7dz8 is created for debugging node... Starting pod/xiyun-20-m-zxpf6-control-plane-1-debug-qhbbr ... To use host binaries, run `chroot /host` Pod IP: 192.168.221.88 If you don't see a command prompt, try pressing enter. sh-4.4# curl 10.131.0.65:8080 ^C sh-4.4# 3. The access can be successful from other master nodes. # oc debug node/xiyun-20-m-zxpf6-control-plane-0 Temporary namespace openshift-debug-kftgw is created for debugging node... Starting pod/xiyun-20-m-zxpf6-control-plane-0-debug-g5f8f ... To use host binaries, run `chroot /host` Pod IP: 192.168.221.89 If you don't see a command prompt, try pressing enter. sh-4.4# curl 10.131.0.65:8080 Hello OpenShift! sh-4.4# exit exit # oc debug node/xiyun-20-m-zxpf6-control-plane-2 Temporary namespace openshift-debug-t29j6 is created for debugging node... Starting pod/xiyun-20-m-zxpf6-control-plane-2-debug-j8n9t ... To use host binaries, run `chroot /host` Pod IP: 192.168.221.87 If you don't see a command prompt, try pressing enter. sh-4.4# curl 10.131.0.65:8080 Hello OpenShift! sh-4.4# exit Exit 4. Created a pod on another work node xiyun-20-m-zxpf6-compute-1 % oc get pods -n test -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES hello-pod 1/1 Running 0 8s 10.128.2.20 xiyun-20-m-zxpf6-compute-1 <none> <none> 5. Try to access this pod from master node xiyun-20-m-zxpf6-control-plane-1, succeed. # oc debug node/xiyun-20-m-zxpf6-control-plane-1 Starting pod/xiyun-20-m-zxpf6-control-plane-1-debug-gfwkz ... To use host binaries, run `chroot /host` Pod IP: 192.168.221.88 If you don't see a command prompt, try pressing enter. sh-4.4# curl 10.128.2.20:8080 Hello OpenShift!
Actual results:
One control plane node cannot access the pod on one worker node. Seen from the reproduce steps
Expected results:
The control plane node should able to access the pod on any worker node
Additional info:
Must gather logs are available at: https://drive.google.com/drive/folders/10NneU_ggcC992ySw1pvLj5qiVCEQk87R
- blocks
-
SDN-4168 Improve ipsec tests
- In Progress
- duplicates
-
OCPBUGS-25312 [OVN][IPSEC EW]Upgrade from 4.14->4.15 failed for Vsphere
- Closed
- is blocked by
-
SDN-4482 Impact statement request for OCPBUGS-22185 [OVN IPsec]One master node cannot access the pod on one worker node
- Closed
- links to