Note the difficulty in interpreting and understanding the text at first reading.

Recent trends in grid scheduling system research represent a movement from centralized based scheduling models to decentralize and hierarchical model due to the lack of dependability and low resource efficiency. Hierarchical models have special notice for some researches, whereas some others believe that it is a type of the centralized model and its disadvantages can also be generalized in the hierarchical model. besides, it can get both advantages of the decentralized and centralized model if we improve some scheduling components in resource management section. One of these components in the hierarchical scheduling system  is the resource discovery and updating management. The basic reason that  we cannot take advantage of earlier research is the assumptions that  underlying centralized and hierarchical models do not hold in grid circumstances and feasibly make weak grid schedules. In this research,  our proposed method is considered as a combinational scheduling model  based on hierarchical and decentralized models. We have utilized  advantages of a decentralized model between local schedulers in which  each local scheduler can schedule a job independently based on found  available nodes if there have been sufficient member nodes in its own  sites (or branches). However, if it could not schedule a job, it devolve  job to up-level schedulers. In this model, meta-scheduler, as the  high-level scheduler, is the root of scheduling system such as other  hierarchical architectures, but it has less intervention in this model.  Obvious that this paper does not concentrate on scheduling methods or  models, but it argues about the procedures of available nodes detection  and using obtained information on the basis of hierarchical  architecture. In practice, our proposed mechanism is separated into two  components. The first component is responsible for online detecting  available nodes and updating their status. The second component is to  identify sufficiently required nodes for the requester (job owner or  consumer) based on input parameters.