This paper addresses the problem of scalability in distributed object systems. It first describes the scaling problem and then uses several examples as discussion points for the participants.

An approach to the secure logon problem in distributed systems managed by a single authority is considered in which central authentication is layered onto existing terminal services. This approach suggests itself when a large installed base of computer systems that do not support central authentication already exists. Work to assess the feasibility of this approach was carried out. The results demonstrate that layering can be used in certain circumstances to provide central authentication services, although, as a result, the concomitant maintenance costs may increase. It was also determined that terminal service features are necessary so that central authentication is easily layered over existing terminal services. Recommendations are made concerning how to structure terminal services in a distributed system to support an integrated central authentication service.>

Recent work examining distributed system security requirements. is critiqued. A notion of trust based on distributed system topology and distributed system node evaluation levels proposed in that work is shown to be deficient. The notion fails to make allowances for the distributed system physical security environment, security factors related to the management of distributed systems by more than one jurisdictive authority, and the interactions that can occur between nodes supporting different mandatory and discretionary security mechanisms.

Recent work examining distributed system security requirements is critiqued. A notion of trust based on distributed system topology and distributed system node evaluation levels proposed in that work is shown to be deficient. The notion fails to make allowances for the distributed system physicat security environment security factors related to the management of distributed systems by more than one jurisdictive authority and interactions that can occur between nodes supporting different mandatory and discretionary security mechanisms.

It is observed that a direct mapping exists between a distributed system's physical configuration and the security threats that can be mounted against interprocess communication in that system. A systematic methodology is presented which implements that mapping for a large class of distributed systems. The methodology includes a model of threats to interprocess communication as well as a model of distributed system security configurations. This methodology is useful in situations where certain major characteristics of the distributed system physical configuration will remain stable over a long time.

Due to advances in integrated circuit technology, multiprocessor systems with more than two Central Processing Units are becoming economically feasible. For this reason, the study of such systems with the aim of developing more efficient hardware and software structures is gaining interest. One type of multiprocessor structure which has recently received some attention is the dynamically microprogrammable multiprocessor system. This paper investigates the problems of efficiently allocating the processors of such a system among the system's workload. A collection of algorithms and scheduling procedures is developed which may be used for processor allocation in dynamically microprogrammable multiprocessor systems with an arbitrary number of processors. Proofs that these algorithms produce configurations which optimise certain measures are given. The algorithms are practical in the sense that the computation required increases linearly with respect to the number of processors to be allocated. Finally, a simulation study has shown that in many circumstances the use of these algorithms can be expected to yield a significant improvement in system performance over an allocation scheme previously proposed.