fix(preconfigured job): Fix for k8s manifest resources limit/request …

[j-sandy]

Fix for #6257

Issue Summary:

Putting limits/request in a custom job through ~/.hal/default/profiles/orca-local.yml, the custom stage breaks with an error as shown below.

About the fix: In order to format above Map suitable for kubectl, "resources.request/limit.memory/cpu" field must take single value. This can be achieved by intercepting the preconfiguredMap object from class com.netflix.spinnaker.orca.clouddriver.pipeline.job.PreconfiguredJobStage and update it with required "resource" field format before passing it further.

[lpcruz]

Hi - running into this exact issue, will start watching this changeset.