In a recent two-layer approach to survivability of networked computing systems migratory autonomous agents have been used as a reactionary mechanism-augmenting a low-level attack recognition scheme. This paper investigates the theoretical implications of determining the agent traversal route as specified by such survivability architecture. It introduces a five-step model that transforms survivability applications into a parameterized graph model that, together with model abstraction and representations, can be the basis for solutions derived from scheduling algorithms. The derivation of agent traversal paths will be transformed into the problem of scheduling jobs with linear precedence order in machine scheduling utilizing weak and strong modes of scheduling.