Skip to Main Content
RT-GRIS, our grid portal, has two purposes. One is effective CPU scavenging and other is supporting a great number of grid node. CPU-scavenging, cycle-scavenging, cycle stealing or shared computing creates a “grid” from the unused resources in a network of participants. It is an important technique for a distribution of grid job and node resource. And, if the grid portal uses fewer resources, those resources are more available to the grid nodes. Eventually the grid portal is supporting a great number of grid nodes. For this reason, the following four kinds of techniques are proposed, first is HDA (hierarchical data access), second is SMF (Standard Message Format), third is GRISP (GRIS Protocol), last is RRM (Rollback Recording Mechanism)/RCM(Reconnecting Mechanism). RT-GRIS has been built and operated at real-fields which are M-Grid and U-Grid using these proposed conception. In this paper, architecture design and implementation of GRIS system are explained. Particularly, performance of RT-GRIS system is evaluated clearly.