Abstract: Methods, apparatus, and processor-readable storage media for management of unit-based virtual accelerator resources are provided herein. An example computer-implemented method includes obtaining multiple accelerator device resource consumption measurements, wherein the measurements represent multiple accelerator device resource types consumed by one or more accelerator devices over a defined temporal interval; computing a composite unit of measurement of accelerator device resource consumption, attributable to the one or more accelerator devices over the defined temporal interval, by normalizing the multiple accelerator device resource consumption measurements using a scaling factor that is based at least in part on one or more static aspects of the one or more accelerator devices; and outputting the composite unit of measurement to at least one user.

Abstract: An apparatus in one embodiment comprises at least one processing device having a processor coupled to a memory configured to implement a first ledger node of a plurality of ledger nodes configured to collectively maintain a distributed ledger. The first ledger node is configured to obtain a set of quality of service metrics for a given workload running on a given cloud service provider, to receive a set of monitoring metrics for the given workload, and to generate a cryptographic block characterizing the set of monitoring metrics, the cryptographic block being entered into the distributed ledger. The first ledger node is also configured to verify whether the given workload meets the set of quality of service metrics based at least in part on the set of monitoring metrics, and to trigger release of resources from a given client device to the given cloud service provider responsive to successful verification.

Abstract: An apparatus in one embodiment comprises at least one processing device having a processor coupled to a memory configured to implement a first ledger node of a plurality of ledger nodes configured to collectively maintain a distributed ledger. The first ledger node is configured to obtain a set of quality of service metrics for a given workload running on a given cloud service provider, to receive a set of monitoring metrics for the given workload, and to generate a cryptographic block characterizing the set of monitoring metrics, the cryptographic block being entered into the distributed ledger. The first ledger node is also configured to verify whether the given workload meets the set of quality of service metrics based at least in part on the set of monitoring metrics, and to trigger release of resources from a given client device to the given cloud service provider responsive to successful verification.

Abstract: A method includes selecting a given workload associated with at least one application type, and analyzing the given workload to determine a set of functional patterns describing resource structures for implementing functionality of the at least one application type. The method also includes determining resource consumption demand profiles for each of the set of functional patterns and determining micro workload distributions for each of the resource consumption demand profiles, a given one of the micro workload distributions comprising a number of micro workloads, each micro workload comprising a normalized unit of resource consumption metrics. The method further includes converting the micro workload distributions for each of the resource consumption demand profiles into a set of resource requirements for running the given workload on an information technology infrastructure, and allocating resources of the information technology infrastructure to the given workload based on the set of resource requirements.

Abstract: A method includes selecting a given workload associated with at least one application type, and analyzing the given workload to determine a set of functional patterns describing resource structures for implementing functionality of the at least one application type. The method also includes determining resource consumption demand profiles for each of the set of functional patterns and determining micro workload distributions for each of the resource consumption demand profiles, a given one of the micro workload distributions comprising a number of micro workloads, each micro workload comprising a normalized unit of resource consumption metrics. The method further includes converting the micro workload distributions for each of the resource consumption demand profiles into a set of resource requirements for running the given workload on an information technology infrastructure, and allocating resources of the information technology infrastructure to the given workload based on the set of resource requirements.

Abstract: Methods, apparatus, and processor-readable storage media for management of unit-based virtual accelerator resources are provided herein. An example computer-implemented method includes obtaining multiple accelerator device resource consumption measurements, wherein the measurements represent multiple accelerator device resource types consumed by one or more accelerator devices over a defined temporal interval; computing a composite unit of measurement of accelerator device resource consumption, attributable to the one or more accelerator devices over the defined temporal interval, by normalizing the multiple accelerator device resource consumption measurements using a scaling factor that is based at least in part on one or more static aspects of the one or more accelerator devices; and outputting the composite unit of measurement to at least one user.

Abstract: An apparatus in one embodiment comprises a processing platform comprising one or more processing devices. The processing platform implements a database configured to store metric and level information for each of a plurality of models, and a metric mapper adapted to interface with the database. The metric mapper is configured to utilize the metric and level information stored in the database to establish an association between a first metric characterizing at least one of a workload and a resource in accordance with a first one of the models and a second metric characterizing at least one of a workload and a resource in accordance with a second one of the models. The established association is illustratively applied to relate service requirements specified at least in part using the first metric of the first model to service capabilities specified at least in part using the second metric of the second model.