[BUG] validation and expansion of Nova filters
While this allows for scaling across compute nodes, always deploying to the node with the most cpu's available, it does not perform properly with different classes of hardware. Overall intent was to have the filter take precedence ignoring the compute node hardware differences, and then simply deploy based off of the number of available vCPU's on a node.
[filter_scheduler]
available_filters = nova.scheduler.filters.all_filters
enabled_filters = ComputeFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter
weight_classes = nova.scheduler.weights.cpu.CPUWeigher
cpu_weight_multiplier = 1.0
#
# Staging for 2024.1 release to base scheduling off instances instead of cpu consumption
#available_filters = nova.scheduler.filters.all_filters
#enabled_filters = ComputeFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter
#weight_classes = nova.scheduler.weights.num_instances.NumInstancesWeigher
#num_instances_weight_multiplier = -1.0