Chapter 4. APIcast operator-based upgrade guide: from 2.9 to 2.10
This section contains information about upgrading APIcast from version 2.9 to 2.10, in an operator-based deployment.
In order to understand the required conditions and procedure, read the entire upgrade guide before applying the listed steps. The upgrade process disrupts the provision of the service until the procedure finishes. Due to this disruption, make sure to have a maintenance window.
4.1. Prerequisites to perform the upgrade
This section describes the required configurations to upgrade APIcast from 2.9 to 2.10 in an operator-based installation.
- APIcast 2.9 previously deployed via the APIcast operator.
- An OpenShift Container Platform (OCP) 4.x cluster with administrator access.
4.2. Upgrading APIcast from 2.9 to 2.10 in an operator-based installation
To upgrade APIcast from version 2.9 to 2.10 in an operator-based deployment:
- Log in to the OCP console using the account with administrator privileges.
- Select the project where the APIcast operator has been deployed.
- Click Operators > Installed Operators.
- In Subscription > Channel, select Red Hat Integration - 3scale APIcast gateway.
Edit the channel of the subscription by selecting the threescale-2.10 channel and save the changes.
- This will start the upgrade process.
- Wait until the upgrade process finishes for APIcast.
Query the pods status on the project:
oc get pods
- Wait until all the new versions are running and ready without errors.
They might have temporary errors during the upgrade process.
NoteTimes can vary from 5-10 minutes approximately. Be sure to keep checking the state of the pods until all of them are running, ready, and without errors.
Check the status of the APIcast objects and get the YAML content by running the following command:
oc get apicast <myapicast> -o yaml
The new annotations with the values should be as follows:
apicast.apps.3scale.net/operator-version: “0.4.0”
After you have performed all the listed steps, APIcast upgrade from 2.9 to 2.10 in an operator-based deployment is now complete.
