CHG0076998 CITZ MCS GOLD - Migrate to OVN Container Network Interface

» View Event Details | Created Mon, 13 Apr 2026 16:00:00 +0000

Scheduled

The current Container Network Interface (CNI) called OpenShift SDN is deprecated and not available in the next version of OpenShift. We need to move to OVNKubernetes CNI to unlock the upgrade path to OpenShift 4.18.

The upgrade requires a rolling restart of all nodes in the cluster twice. API nodes and Infra nodes are highly available and will fail over services as needed.

As the node drains are managed by OpenShift operator automation, Platform Ops has no way to control the speed of drains.

We may continue to see slow or delayed storage operations like mounting PVCs to pods. Pods drained from one node onto another node may see up to 45 min delay until it starts. The best solution is to have multiple replicas so that a single pod failing to start does not cause application downtime.

CHG0076035 STMS Calgary Data Centre Removal of Extension switch from MCS Load balancer

» View Event Details | Created Sun, 19 Apr 2026 11:00:00 +0000

Scheduled

The Extension switch connecting the MCS Load balancer to the Core switches will be bypassed.

More details at the TIB https://ociomysc.service-now.com/sp?id=kb_article&table=u_kb_template_technical_information_bulletin&sys_id=e51bc58bfb3b369492f6fa686eefdc00&view=ess

There will be an outage of up to 30 minutes to all Gold DR OpenShift traffic.

Should a rollback be necessary, there will be a second outage of up to 30 minutes during the change window.

Clients are encouraged to test services and validate that applications are functioning before the change window and after the activity has completed.

CHG0076999 CITZ MCS SILVER - Migrate to OVN Container Network Interface

» View Event Details | Created Mon, 20 Apr 2026 16:00:00 +0000

Scheduled

The current Container Network Interface (CNI) called OpenShift SDN is deprecated and not available in the next version of OpenShift. We need to move to OVNKubernetes CNI to unlock the upgrade path to OpenShift 4.18.

The upgrade requires a rolling restart of all nodes in the cluster twice. API nodes and Infra nodes are highly available and will fail over services as needed.

As the node drains are managed by OpenShift operator automation, Platform Ops has no way to control the speed of drains.

We may continue to see slow or delayed storage operations like mounting PVCs to pods. Pods drained from one node onto another node may see up to 45 min delay until it starts. The best solution is to have multiple replicas so that a single pod failing to start does not cause application downtime.