Upgrade your Airy Core instance
Upgrade an existing installation of Airy Core.
#
Upgrade your CLIIn order to upgrade your Airy Core instance, first you need to upgrade your Airy CLI, depending on your installation method.
You can check the current version of your Airy CLI with the airy version
command:
#
Upgrade your Airy Core instancewarning
The upgrade of your Airy Core cluster will lead to downtime. Usually it takes seconds, but in case there is a migration or a reset of some of the streaming apps, this process might take longer, depending on the amount of data you have.
Before you proceed, refer to the Release documentation for any upgrade notes or steps.
The upgrade process will not delete any of the persistent data that is kept inside the Kafka cluster. Use the airy upgrade
command to perform the upgrade of your Airy Core instance. Run airy upgrade
to return information about your current Airy Core version and the latest version available. You will be prompted to proceed with the upgrade (omit this prompt by using the --approve
flag).
note
You need to be inside an Airy Core workspace directory to run the command.
You can overwrite the path by specifying the --workspace
flag, for example:
airy upgrade --workspace ~/.airy/production
.
The upgrade will continue until complete. You will receive a notification to confirm that the upgrade was successful. The version will be updated inside the Kubernetes cluster and in your local airy.yaml
file.
#
Upgrade using HelmIf you used Helm to deploy Airy Core
, you can upgrade with the helm upgrade
command:
or to upgrade to a specific version:
#
Cleanup the upgradeThe upgrade will create few additional resources (Kubernetes jobs and configMaps) which are not deleted automatically, so that there is a better insight of what happened during the upgrade. To cleanup those resources, run:
#
RollbackIn case the upgrade fails, first you can inspect the reason for the failure by running this command:
After that, you can rollback to your previous Airy Core version. Clean-up the upgrade resources as instructed in the previous section and run:
#
TroubleshootingIf you need further help, refer to our Troubleshooting section.