Browse Prior Art Database

A Method and System for Enabling a Hypervisor to Coalesce Network Interrupts for Each Partition in a Micro-Partitioned Environment

IP.com Disclosure Number: IPCOM000198064D
Publication Date: 2010-Jul-24
Document File: 1 page(s) / 20K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method and system is disclosed for enabling a hypervisor to coalesce network interrupts for each partition in a micro-partitioned environment. Subsequently, the network interrupts for each partition are periodically serviced based on a user defined configuration.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 76% of the total text.

Page 1 of 1

A Method and System for Enabling a Hypervisor to Coalesce Network Interrupts for Each Partition in a Micro-Partitioned Environment

Disclosed is a method and system for enabling a hypervisor to coalesce network interrupts for each partition in a micro-partitioned environment. Subsequently, the network interrupts for each partition are periodically serviced based on a user defined configuration.

In accordance with the method and system disclosed herein, a hypervisor is configured to maintain an interrupts queue for each partition of the micro-partitioned environment for implementing an interrupt coalescing mechanism. Accordingly, when a hypervisor is invoked upon receiving of an interrupt, the hypervisor determines a target partition of the interrupt and stores the interrupt in an interrupts queue associated with the target partition before servicing the interrupt.

In an embodiment, the hypervisor may wait for a predefined time interval before servicing the interrupts stored in the interrupts queue. Accordingly, the hypervisor interrupts (dispatch) the target partition once every predefined time interval. The predefined time interval is defined so as to make sure that the interrupts stored in the interrupts queue do not have high latencies penalties.

In another embodiment, the hypervisor may wait for a predefined number of interrupts to be stored in the interrupts queue before servicing the interrupts. Accordingly, the hypervisor interrupts (dispatch) the target pa...