Upgrades and Rollbacks

Upgrading to a later version of any of the components that run as part of the network can be done by running the same command (assuming an updated configuration is provided).

NOTE: If upgrading from FPX helm chart before version 2.1.4, there is a manual step required. Before upgrading, you must manually switch any existing deploy services from type Nodeport to ClusterIP (ensuring not only the spec.type is updated, but the spec.ports.nodeports is removed as well). This would prevent which would otherwise be a conflict during the update process.

To do so, run the following:

kubectl edit svc fpx-authapi

kubectl edit svc fpx-authserver

$ helm upgrade --install fpx identos-helm/fpxcore -f configuration.yaml --version v2.1.4

The upgrade process is complete when you run the following command and all deployments are ready, up-to-date and available:

$ kubectl get deployments
NAME                 READY   UP-TO-DATE   AVAILABLE   AGE
fpx-authapi          1/1     1            1           1m42s
fpx-authserver       1/1     1            1           1m42s

Rollback

To rollback to previous version of FPX, execute the following command:

$ helm rollback [revision] fpx

Refer to https://helm.sh/docs/helm/helm_rollback/ for more details related to rollback of Helm charts.

Note: Some rollbacks may require database changes. Please refer to the version specific release notes for more details related to DB updates/rollbacks.

Rollback​

Rollback