-
Bug
-
Resolution: Can't Do
-
Normal
-
None
-
4.12.z, 4.14.z
-
Important
-
No
-
False
-
In an OpenShift cluster with an OpenShiftSDN network plugin with egressIP and NMstate operator configured, some conditions when the egress IP is moved from one node to another node, the NMstate operator is not realising the egressIP from the old node.
New node(where the EgressIP has switched):
[root@x18bospsbas1a# ssh core@x18sospspe602a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
inet 10.52.48.125/24 brd 10.52.48.255 scope global secondary ens192:eip
Old node(where the EgressIP was present):
[root@x18bospsbas1a# ssh core@x18sospspe603a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
inet 10.52.48.125/24 brd 10.52.48.255 scope global secondary noprefixroute ens192
NOTE: 10.52.48.125 is the egressIP, and on the hostsubnet we are able to see that openshift level the node x18sospspe602a.ocp6.uat.dbs.com that holds the real egressIp address.
[ocpadm1@x18bospsbas1a$ oc get hostsubnets | egrep "tmup|10.52.48.125"
x18sospspe602a.ocp6.uat.dbs.com x18sospspe602a.ocp6.uat.dbs.com 10.52.48.2 192.168.8.0/25 ["10.52.48.0/24"]
How reproducible: NA
Steps to Reproduce:
1. Build openshift-cluster with the SDN networking.
2. Install the NMstate operator and configure a new DNS server information on the nodes (master +worker).
3. Add an egress IP address to the namespace and map it to the node.
4. Make the egressIp to switch from one node to another node.
5. check the EgressIP on the both the nodes on it's network interface.
Actual results:
[root@x18bospsbas1a# ssh core@x18sospspe601a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
[root@x18bospsbas1a# ssh core@x18sospspe602a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
inet 10.52.48.125/24 brd 10.52.48.255 scope global secondary ens192:eip
[root@x18bospsbas1a# ssh core@x18sospspe603a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
inet 10.52.48.125/24 brd 10.52.48.255 scope global secondary noprefixroute ens192
[root@x18bospsbas1a#
Expected results:
[root@x18bospsbas1a# ssh core@x18sospspe601a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
[root@x18bospsbas1a# ssh core@x18sospspe602a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
inet 10.52.48.125/24 brd 10.52.48.255 scope global secondary ens192:eip
[root@x18bospsbas1a# ssh core@x18sospspe603a.ocp6.uat.dbs.com "ip a show dev ens192 | grep 10.52.48.125"
[root@x18bospsbas1a#
Affected Platforms: Openshift Container Platfrom.
Is it an customer issue / SD