The effects of cache memories on the performance of interconnect controllers (ICs) are analyzed using trace-driven simulation. Simulation results show that the controllers which interconnect hosts to LANs have a higher hit ratio than those that interconnect network devices to LANs. The impact of cache sizes, set associativity, and line sizes on cache performance is also investigated. A significant observation is that, although increasing the sizes can result in a higher hit ratio, it can also considerably increase traffic to main memory, thereby degrading overall system performance. A method of determining an optimal line size that produces the best overall system performance is therefore needed. A simple analytical model for determining the optimal line size as a function of cache size is presented.>

Since each of the levels in a hierarchical system could have various characteristics, different fault-tolerant schemes could be appropriate at different levels. A stochastic Petri net (SPN) is used to investigate various fault-tolerant schemes in this context. The basic SPN is augmented by parameterized subnet primitives to model the fault-tolerant schemes. Both centralized and distributed fault-tolerant schemes are considered. The two schemes are investigated by considering the individual levels in a hierarchical system independently. In the case of distributed fault tolerance, two different checkpointing strategies are considered. The first scheme is called the arbitrary checkpointing strategy. Each process in this scheme does its checkpointing independently; thus, the domino effect may occur. The second scheme is called the planned strategy. Here, process checkpointing is constrained to ensure no domino effect. The results show that, under certain conditions, an arbitrary checkpointing strategy can perform better than a planned strategy. The effect of integration on the fault-tolerant strategies of the various levels of a hierarchy are studied.>

Analytical models are presented that use Petri nets for fault-tolerant schemes used in distributed systems. These models are used in the quantitative evaluation and selection of good fault-tolerant schemes for specific system configurations. Several different fault-tolerant schemes that can be modeled using Petri nets are discussed in detail. These schemes include rollback recovery with checkpointing, recovery blocks, N-version programming, and conversations. After a brief review of Petri net models, extension of the Petri net models to incorporate fault-tolerant schemes is considered. A methodology for evaluating a fault-tolerant scheme for a specific system configuration and the steps involved in building a Petri net model of a fault-tolerant system are described. The subnet primitives involved in building these models are identified and an algorithm for building the models automatically is described. Examples illustrating this extended Petri net model are discussed and numerical results are presented to show the applicability of the models.>