1. Upgrading Koios Automation Platform

Refer to Starting, Stopping, and Restarting Koios Automation Platform for detail.

1.1. Upgrade Planning

This section covers changes that you should keep in mind as you attempt to upgrade your Koios Automation Platform instance.

  • Clustered upgrades require special attention to instance and instance groups prior to starting the upgrade. See Clustering for details.

  • Back up the deployment before upgrading. See Backup and Restore.

For the upgrade procedure itself, see Install.

1.2. Reissued releases

A release with a known defect is reissued with the same number and a new catalog digest. A cluster that already runs the first issue does not upgrade by itself, because the version does not change:

  • With OLM (OpenShift or Kubernetes): point the koios-operators CatalogSource at the new catalog digest and delete the CSV of that version in the namespace of the operators (for example koios-operator.v2026.11.6-cluster-scoped, or koios-operator.v2026.11.6 in the namespaced channel). OLM installs it again from the new catalog. If the Subscription stays in UpgradePending, delete it and create it again with the same specification. The platform and its data are not affected; the operators then roll out the reissued images.

  • Without OLM: the tag of that release in koios-installer is moved to the reissue. A checkout made before keeps the first issue: update the tag explicitly (git fetch --tags --force), check that it points to the commit named in the release notes, and apply kustomize/operators again from it.

The release notes of each reissue name the withdrawn catalog and the CSV to delete.