The following timeseries only provide an id field which can be a rack or silo or project id (in the future, we may add a fleet-level rollup as well when there are multiple racks):
collection_target:cpus_provisioned
collection_target:ram_provisioned
collection_target:virtual_disk_space_provisioned
It'll be very useful to capture the collection_type from the metrics source (which I believe currently is the virtual_provisioning_collection database table) so that users can graph these usage metrics separately based on the aggregation level.
The following timeseries only provide an
idfield which can be a rack or silo or project id (in the future, we may add a fleet-level rollup as well when there are multiple racks):It'll be very useful to capture the
collection_typefrom the metrics source (which I believe currently is thevirtual_provisioning_collectiondatabase table) so that users can graph these usage metrics separately based on the aggregation level.