Hybrid Checkpointing for Parallel Applications in Cluster Federations
Résumé
Cluster federations are very useful for applications like large scale code coupling. Faults may appear very frequently, so we want to use checkpoints to be able to restart applications. To take into account the constraints introduced by clusters federation architecture, we propose a hierarchical checkpointing protocol. It uses synchronization inside clusters but only quasi-synchronous methods between clusters. Our protocol has been evaluate by simulation and fits well for applications that can be divided in modules with a lot of communications inside modules but few between them.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...