International Business Machines Corporation

Italcementi

Algosource Technologies

Christophe Lombard

Khalid Filali-Adib

John Mark McConaughy

David W Sheffield

Sungjin Yook

Hua Chai

Jun He

Stephen Bailey Peckham

Frederic Barrat

Philippe Bergheaud

Laurent Dufour

Luke Mathew Browning

Donald Ray Stence

Guy Menanteau

William G Tuel, Jr

Christophe Fagiano

Richard J Coppinger

Gary J Mincher

Fabio Corazza

Christophe Pierre Francois Quintard

Olivier Lepine

Jérôme Leuranguer

François Leborgne

A method of performing a checkpoint on a set of connected processors and memories comprises the steps of creating one or more statefiles for one or more of the processors, querying available processing and/or memory resources, allocating data from one or more statefiles to the available resources, compressing the allocated data, storing the compressed data, and repeating the querying, allocating, compressing and storing steps until all of the statefile(s) are compressed and stored.

Checkpoint and restart

Mechanisms are provided for performing a live update of an operating system. The mechanisms receive an update to an operating system and clone a root volume group associated with an operating system instance executing in a first logical partition of the data processing system to generate a cloned root volume group. The mechanisms apply the update to the cloned root volume group to generate an updated and cloned root volume group and boot a second logical partition of the data processing system using the updated and cloned root volume group. Moreover, the mechanisms migrate application instances and application data to the second logical partition.

Live operating system update mechanisms

Mechanisms are provided for performing a live update of an operating system. The mechanisms receive an update to an operating system and clone a root volume group associated with an operating system instance executing in a first logical partition of the data processing system to generate a cloned root volume group. The mechanisms apply the update to the cloned root volume group to generate an updated and cloned root volume group and boot a second logical partition of the data processing system using the updated and cloned root volume group. Moreover, the mechanisms mirror the original root volume group associated with an operating system instance executing in a first logical partition and import the mirrored root volume group into the second logical partition. The mechanisms migrate application instances to the second logical partition by restarting the application instances in the second logical partition using the mirrored root volume group.

A hierarchical fanout connectivity infrastructure is built and used to start a parallel application within a parallel computing environment. The connectivity infrastructure is passed to a checkpoint library, which employs the infrastructure and a defined sequence of events, to perform checkpoint, restart and/or migration operations on the parallel application.

Fanout connectivity structure for use in facilitating processing within a parallel computing environment

Growing community of inventors

Saint Paul de Vence, France

Christophe Lombard