Abstract: In some examples, a system receives a first unit of work to be scheduled in the system that includes a plurality of collections of processing units to execute units of work, where each respective collection of processing units of the plurality of collections of processing units is associated with a corresponding scheduling queue. The system selects, for the first unit of work according to a first criterion, candidate collections from among the plurality of collections of processing units, and enqueues the first unit of work in a schedule queue associated with a selected collection of processing units that is selected, according to a selection criterion, from among the candidate collections.

Abstract: In some examples, a system receives a first unit of work to be scheduled in the system that includes a plurality of collections of processing units to execute units of work, where each respective collection of processing units of the plurality of collections of processing units is associated with a corresponding scheduling queue. The system selects, for the first unit of work according to a first criterion, candidate collections from among the plurality of collections of processing units, and enqueues the first unit of work in a schedule queue associated with a selected collection of processing units that is selected, according to a selection criterion, from among the candidate collections.

Abstract: A computer implemented method includes: storing identifiers of cached data blocks stored in a ring structure comprising a plurality of queues, where each queue is assigned a different priority level based on a base pointer, and where the plurality of queues includes a subset of queues associated with an insertion range of priority levels; during a plurality of partial rotations of the ring structure, inserting new identifiers of data blocks associated with a plurality of processing engines into the subset of queues, where each partial rotation shifts the base pointer to a next higher queue; and maintaining, for multiple partial rotations, an association between a first processing engine and a first queue of the subset of queues until the first queue is shifted out of the insertion range of priority levels.

Abstract: A computer implemented method includes: storing identifiers of cached data blocks stored in a ring structure comprising a plurality of queues, where each queue is assigned a different priority level based on a base pointer, and where the plurality of queues includes a subset of queues associated with an insertion range of priority levels; during a plurality of partial rotations of the ring structure, inserting new identifiers of data blocks associated with a plurality of processing engines into the subset of queues, where each partial rotation shifts the base pointer to a next higher queue; and maintaining, for multiple partial rotations, an association between a first processing engine and a first queue of the subset of queues until the first queue is shifted out of the insertion range of priority levels.