The OpenShift Pro cluster, pro-us-east-1, is scheduled to upgrade to version 3.9.31 starting at 16:00 UTC on Jul 19, 2018 and ending at approximately 00:00 UTC on Jul 20, 2018. The upgrade process is engineered to disrupt applications as little as possible, but brief outages may occur during the upgrade. For short periods of time, developers may be unable to create new applications or interact with existing applications. In addition to the OpenShift version upgrade, the following imagestreams will also be updated:
To disable automatic upgrade of these applications, modify the imageChangeTrigger in the application's DeployConfig or BuildConfig.https://docs.openshift.com/container-platform/3.9/dev_guide/deployments/basic_deployment_operations.html#image-change-trigger
How does the upgrade work?
During the 3.9 in-place upgrade, existing nodes will be evacuated, which requires pods to be stopped and moved to other nodes in the cluster.
What should you do?
You can minimize the impact on your applications by scaling your services to more than one(1) pod. In general, for applications to be able to continue to service clients, they should be scaled. Some pod workloads are not appropriate for scaling, such as a single-instance, non-replicated database using a persistent volume claim. In this situation, a deployment strategy of 'recreate' will ensure the pod is restarted after migration, although a brief outage will be experienced.
For more information, refer to the following guide:https://blog.openshift.com/deploying-highly-available-applications-openshift-kubernetes/