PSA: An issue was found in Harvester v1.4.2 where high CPU usage may occur after an upgrade because of the

backup-target

setting's new

refreshIntervalInSeconds

field. If the field is left empty or is set to 0, Harvester constantly refreshes the backup target, resulting in high CPU usage. To fix the issue, update the value of

refreshIntervalInSeconds

to a larger number (for example, 60) using the command

kubectl edit setting backup-target

. You can also update the value before starting the upgrade to prevent the issue from occurring. Example:

value: '{"type":"nfs","endpoint":"<nfs://longhorn-test-nfs-svc.default>:/opt/backupstore", "refreshIntervalInSeconds": 60}'

Ref: https://github.com/harvester/harvester/issues/7885 This information has been added to the release notes at https://github.com/harvester/harvester/releases/tag/v1.4.2 and the documentation at https://docs.harvesterhci.io/v1.4/upgrade/v1-4-1-to-v1-4-2/