Removed:
- Metrics and pod logs collection. Crust gather collects logs for all
resources.
Fixed:
- MachineDeployment checks for running machines. MachineSets are picked
at random, as they are indistinguishable based on labels, and belong
to the same MachineDeployment. This causes flakes as old MachineSet is
expected to scale accordingly, while the new one performed it instead.
- Increased ClusterClass apply timeouts. CAPD webhooks may take longer
to stand up.
Signed-off-by: Danil-Grigorev <danil.grigorev@suse.com>
Andrea Mazzotti
Alexandr Demicev
Furkat Gofurov
Danil-Grigorev
Carlos Salas