Description of problem: Dangling lr-route with ICNI 2.0 even after deletion of ICNI router namespace (including the pod with the required annotations) and the application namespace I pushed an unrelated MCP modification and MCP reconfigured and restarted my servers. After a few days, I came back to my environment and my ICNI 2.0 setup had stopped working. I tried to reprovision the application by completely deleting the router and application namespace but to no avail. The cluster is on 4.7.13: ~~~ [root@openshift-jumpserver-0 ~]# oc get clusterversion NAME VERSION AVAILABLE PROGRESSING SINCE STATUS version 4.7.13 True False 11d Cluster version is 4.7.13 ~~~ After I deleted my ICNI 2.0 router and the application (both the router and the application namespace were completely removed), I still saw some dangling routes: ~~~ [root@openshift-jumpserver-0 ~]# oc exec -n openshift-ovn-kubernetes -it $POD -- ovn-nbctl lr-route-list GR_openshift-worker-1 Defaulting container name to ovn-controller. Use 'oc describe pod/ovnkube-node-7s2kh -n openshift-ovn-kubernetes' to see all of the containers in this pod. IPv4 Routes 172.25.2.18 192.168.123.120 src-ip ecmp-symmetric-reply # <--- here 172.25.3.171 192.168.123.120 src-ip ecmp-symmetric-reply # <--- here 172.24.0.0/14 100.64.0.1 dst-ip 0.0.0.0/0 192.168.123.1 dst-ip rtoe-GR_openshift-worker-1 ~~~ The application pods did not exist any more: ~~~ [root@openshift-jumpserver-0 ~]# oc get pods -A -o wide | grep 172.25.2.18 [root@openshift-jumpserver-0 ~]# oc get pods -A -o wide | grep 172.25.3.171 openshift-monitoring thanos-querier-7668d5d5f8-pvhqk 5/5 Running 0 5d20h 172.25.3.171 openshift-worker-1 <none> <none> ~~~ Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: Further details in my private comment.
*** This bug has been marked as a duplicate of bug 1978797 ***