Grid resources having heterogeneous architectures, being geographically distributed and interconnected via unreliable network media, are at the risk of failure. Grid environment consists of unreliable resources; therefore, fault tolerant mechanisms can not be ignored. Some scientific jobs require long commitments of grid resources whose failures may not be overlooked. We need a flexible management of these failures by considering the failure of fault manager itself. In this paper we propose the concept of distributed management of failures without engaging the resources for this particular task exclusively. Resources performing the fault management may also participate in serving the long running user jobs. Each sub-job of the main user job is inspected by an individual resource. In case of failure inspector resource takes over in place of inspected resource. Contributions of this paper are: elimination of single point of failure and proposed concept?s ability to be integrated with variety of grid middleware.