For the Kubernetes function worker runtime, allow customization of the pulsar function "jobName"

[csthomas1]

Is your enhancement request related to a problem? Please describe.
Currently (as of 2.6.x), the Pulsar KubernetesRuntime class hardcodes the jobName (the name assigned to the StatefulSet used to create the function pods) to the format "pf-[tenant]-[namespace]-[function][-optional 8 char hash]." While the intent of this name format was no doubt both to provide a human readable name for the k8s objects and ensure uniqueness within k8s, we've found it -- when combined with the 55 character size restriction imposed by KubernetesRuntime -- to be unnecessarily limiting. In our environment, we ensure that Pulsar functions under a particular Pulsar tenant deploy into a kubernetes namespace dedicated to that tenant; hence, for us the [tenant] portion of the function name is redundant. Further, the "pf-" prefix is unnecessary, as we're able to distinguish the function pods from other pods based on the function name alone. These issues may seem minor, but they consume precious characters against the 55 character max!

Describe the solution you'd like
Allow us to set the k8s jobName as a custom runtime parameter (similar to jobNamespace). Attach labels to the k8s StatefulSet to separately identify the corresponding Pulsar tenant, namespace, and function so as to enable the Function worker to locate and clean up the StatefulSet when requested.

Describe alternatives you've considered
We briefly considered writing an implementation of the KubernetesManifestCustomizer interface that could replace the generated job name with a custom version, but then observed that the KubernetesRuntime class would reset it to the generated value.