Description of problem: Node IP addr is updated but SDN master failed to update corresponding HostSubnet record. Version-Release number of selected component (if applicable): oc v3.9.0-alpha.3+57f5a29-160-dirty kubernetes v1.9.1+a0ce1bc657 How reproducible: Always Steps to Reproduce: 1. Assume node has IP1 and IP2 ip addresses 2. Set hostname as IP1 and start openshift node and openshift master service 3. Update node addr to IP2 by using nodeIP config param in node-config.yaml 4. Restart openshift node service Actual results: Openshift node service fails with unable to fetch local subnet. Expected results: Openshift node should start with new IP addr Additional info:
This is side effect of https://bugzilla.redhat.com/show_bug.cgi?id=1438402 When node IP is updated, we must update host IP in HostSubnet record otherwise updated node will be out of sync with other nodes in the cluster. To avoid node flip flop in case of multiple node IPs, we should always specify nodeIP config option in node-config.yaml
Fixed by https://github.com/openshift/origin/pull/18281
Commit pushed to master at https://github.com/openshift/origin https://github.com/openshift/origin/commit/48a01225a8f693902904b30d72d51e390f8ab708 Merge pull request #18281 from pravisankar/fix-master-nodeevent Automatic merge from submit-queue. Bug 1538389 - Allow node IP change to update Host IP in HostSubnet resource - Node status addresses may have both old and new IP address and by validating against these addrs we may not be able to update HostSubnet with new node IP. Openshift node service waits for its HostSubnet resource with new Host IP and eventually fails. - This change fixes the above issue by reverting the change for node flip/flop. Recommended/correct way to handle node flip/flop case is to specify nodeIP config option in node-config.yaml
verified in openshift v3.9.0-0.38.0 and issue has been fixed. the hostsubnet is updated after updating nodeIP in node-config.yaml and restarting node service.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2018:0489