International Business Machines Corporation

Hongjie Yang

Steven S Shultz

Stephen G Wilkins

Matthew J Rosato

R Scott Coyle

In an exemplary embodiment of this disclosure, a system may include a first hypervisor, a second hypervisor, and a shared memory set. The first hypervisor and the second hypervisor run over a single logical partition. The shared memory set is accessible by both the first hypervisor and the second hypervisor, and the first hypervisor is configured to communicate with the second hypervisor by writing to the shared memory. The second hypervisor may provide support for hardware that is not supported by the first hypervisor, which hardware may be accessed by the first hypervisor through communications facilitated by use of the shared memory.

Systems and methods to offload hardware support using a hypervisor subpartition

Embodiments relate to a method for encapsulating a virtual server in a hypervisor subpartition of a single logical partition to provide additional hardware support and advanced virtualization functions. An aspect includes running a first hypervisor and one or more additional hypervisors over a single logical partition. A request to manage a virtual server is received at the first hypervisor. The request is then analyzed to determine what resources are necessary to manage the virtual server. Based on the analyzed request, one of the first hypervisor or the one or more additional hypervisors is selected to complete the request.

Offloading service requests to a second guest hypervisor in a logical partition shared by a plurality of guest hypervisors

A request server automatically detects, in response to initialization of the request server, if there are previous generations of requests outstanding for the request server, and if so, immunizes itself against those requests. The request server starts one or more threads that are designed to gather messages associated with requests from previous incarnations of the server and handles those messages without affecting the requestors. While the request server is immunized from earlier generations of requests, the request server is capable of processing new requests.

Growing community of inventors

Vestal, NY, United States of America

Hongjie Yang