Controller: Temporarily disable HPAs to fix pod create/delete loop
Temporarily disables the setupAutoscaling
method to address an issue
where the HPAs attempt to create replicas, but the Operator then tries
to remove them, meaning the Deployments never end up showing as
Ready/Available.
This is a temporary fix until we do one of the following:
- Deploy HPAs from the Chart and not the Operator, and support disabling the HPAs via values in the Chart so we can control which ones are created if enabled
- Understand how to update the Operator to accept the replicas created by the HPAs
Note that if HPAs are to be used at all, we need to install the metrics server with our automation when creating the cluster, as it seems the metrics server is not running with our existing installation.
Closes #90 (closed)
Edited by Mitchell Nielsen