NUMA Scheduling Using Inter-vCPU Memory Access Estimation

摘要

In a system having non-uniform memory access architecture, with a plurality of nodes, memory access by entities such as virtual CPUs is estimated by invalidating a selected sub-set of memory units, and then detecting and compiling access statistics, for example by counting the page faults that arise when any virtual CPU accesses an invalidated memory unit. The entities, or pairs of entities, may then be migrated or otherwise co-located on the node for which they have greatest memory locality.

主权项

1. A method for managing memory in a system, said system including a plurality of software entities that each access the memory and having a non-uniform memory access architecture (NUMA) with a plurality of nodes, the method comprising:
selecting a sample set of memory units; invalidating the sample set of memory units and detecting accesses by any of the software entities to the invalidated memory units; computing a metric as a function of a proportion of accessed memory units relative to the sample set of memory units; and associating at least one of the software entities with one of the nodes based at least in part on the corresponding metric so as to increase memory locality of the associated software entity.