E.D. Jensen

An Optimal Real-Time Scheduling Algorithm for Multiprocessors

We present an optimal real-time scheduling algorithm for multiprocessors

On recent advances in time/utility function real-time scheduling and resource management

one that satisfies all task deadlines, when the total utilization demand does not exceed the utilization capacity of the processors. The algorithm called LLREF, is designed based on a novel abstraction for reasoning about task execution behavior on multiprocessors: the time and local execution time domain plane (or T-L plane). LLREF is based on the fluid scheduling model and the fairness notion, and uses the T-L plane to describe fluid schedules without using time quanta, unlike the optimal Pfair algorithm (which uses time quanta). We show that scheduling for multiprocessors can be viewed as repeatedly occurring T-L planes, and feasibly scheduling on a single T-L plane results in the optimal schedule. We analytically establish the optimality of LLREF. Further, we establish that the algorithm has bounded overhead, and this bound is independent of time quanta (unlike Pfair). Our simulation results validate our analysis on the algorithm overhead

A utility accrual scheduling algorithm for real-time activities with mutual exclusion resource constraints

We argue that the key underpinning of the current state-of-the real-time practice - the priority artifact - and that of the current state-of-the real-time art - deadline-based timeliness optimality - are entirely inadequate for specifying timeliness objectives, for reasoning about timeliness behavior, and for performing resource management that can dependably satisfy timeliness objectives in many dynamic real-time systems. We argue that time/utility functions and the utility accrual scheduling paradigm provide a more generalized, adaptive, and flexible approach. Recent research in the utility accrual paradigm has significantly advanced the state-of-the-art of that paradigm. We survey these advances.

Utility accrual real-time scheduling under variable cost functions

This paper presents a uni-processor real-time scheduling algorithm called the generic utility scheduling algorithm (which we refer to simply as GUS). GUS solves a previously open real-time scheduling problem-scheduling application activities that have time constraints specified using arbitrarily shaped time/utility functions and have mutual exclusion resource constraints. A time/ utility function are a time constraint specification that describes an activitys utility to the system as a function of that activitys completion time. Given such time and resource constraints, we consider the scheduling objective of maximizing the total utility that is accrued by the completion of all activities. Since this problem is NP-hard, GUS heuristically computes schedules with a polynomial-time cost of O(n/sup 3/) at each scheduling event, where n is the number of activities in the ready queue. We evaluate the performance of GUS through simulation and by an actual implementation on a real-time POSIX operating system. Our simulation studies and implementation measurements reveal that GUS performs close to, if not better than, the existing algorithms for the cases that they apply. Furthermore, we analytically establish several properties of GUS.

Response time analysis of software transactional memory-based distributed real-time systems

We present a utility accrual real-time scheduling algorithm called CIC-VCUA for tasks whose execution times are functions of their starting times (and, potentially, other factors). We model such variable execution times using variable cost functions (or VCFs). The algorithm considers application activities that are subject to time/utility function time constraints, execution times described using VCFs, and mutual exclusion constraints on concurrent sharing of non-CPU resources. We consider the twofold scheduling objective of 1) assuring that the maximum interval between any two consecutive, successful completions of job instances in an activity must not exceed the activity period (an application-specific objective) and 2) maximizing the systems total accrued utility while satisfying mutual exclusion resource constraints. Since the scheduling problem is intractable, CIC-VCUA is a polynomial-time heuristic algorithm. The algorithm statically computes worst-case task sojourn times, dynamically selects tasks for execution based on their potential utility density, and completes tasks at specific times. We establish that CIC-VCUA achieves optimal timeliness during underloads, and tightly upper bounds inter and intratask completion times. Our simulation experiments confirm the algorithms effectiveness and superiority

Recovering from Distributable Thread Failures with Assured Timeliness in Real-Time Distributed Systems

We consider distributed real-time systems where concurrency control is managed using software transactional memory (or STM). For such a method we propose an algorithm to compute an upper bound on the response time. We compare the result of the proposed algorithm to a simulation of the system being studied in order to determine its efficacy. The results of our study indicate that it is possible to provide timeliness assurances for systems programmed using STM.

On bounding response times under software transactional memory in distributed multiprocessor real-time systems

We consider the problem of recovering from failures of distributable threads with assured timeliness. When a node hosting a portion of a distributable thread fails, it causes orphans - i.e., thread segments that are disconnected from the threads root. We consider a termination model for recovering from such failures, where the orphans must be detected and aborted, and failure-exception notification must be delivered to the farthest, contiguous surviving thread segment for resuming thread execution. We present a realtime scheduling algorithm called AUA, and a distributable thread integrity protocol called TP-TR. We show that AUA and TP-TR bound the orphan cleanup and recovery time, thereby bounding thread starvation durations, and maximize the total thread accrued timeliness utility. We implement AUA and TP-TR in a real-time middleware that supports distributable threads. Our experimental studies with the implementation validate the algorithm/protocols time-bounded recovery property and confirm their effectiveness

Scheduling distributable real-time threads in Tempus middleware

We consider multiprocessor distributed real-time systems where concurrency control is managed using software transactional memory (or STM). For such a system, we propose an algorithm to compute an upper bound on the response time.The proposed algorithm can be used to study the behavior of systems where node crash failures are possible. We compare the result of the proposed algorithm to a simulation of the system being studied in order to determine its efficacy. The results of our study indicate that it is possible to provide timeliness guarantees for multiprocessor distributed systems programmed using STM.

Utility Accrual Real-Time Scheduling under Variable Cost Functions

This paper presents the Tempus real-time middleware, which supports real-time CORBA 2.0s distributable threads (DTs) as an end-to-end programming abstraction for distributed real-time systems. DTs in Tempus can have time constraints, including time/utility functions (TUFs), can have resource constraints, including mutual exclusion, and can be scheduled according to utility accrual (UA) disciplines. Tempus propagates the scheduling parameters of DTs as they transit objects and hence perhaps node boundaries. Node-local instances of a UA scheduling algorithm use the propagated parameters to construct local schedules and resolve resource dependencies for local timeliness optimization, toward approximate, system-wide timeliness optimality. Tempus uses an application-level scheduling framework for node-local TUF/UA scheduling on real-time POSIX-compliant operating systems. Our experimental measurements demonstrate the effectiveness of the middleware in scheduling DTs.

A space-optimal wait-free real-time synchronization protocol

We present a utility accrual real-time scheduling algorithm called CIC-VCUA for tasks whose execution times are functions of their starting times (and, potentially, other factors). We model such variable execution times using variable cost functions (or VCFs). The algorithm considers application activities that are subject to time/utility function time constraints, execution times described using VCFs, and mutual exclusion constraints on concurrent sharing of non-CPU resources. We consider the twofold scheduling objective of 1) assuring that the maximum interval between any two consecutive, successful completions of job instances in an activity must not exceed the activity period (an application-specific objective) and 2) maximizing the systems total accrued utility while satisfying mutual exclusion resource constraints. Since the scheduling problem is intractable, CIC-VCUA is a polynomial-time heuristic algorithm. The algorithm statically computes worst-case task sojourn times, dynamically selects tasks for execution based on their potential utility density, and completes tasks at specific times. We establish that CIC-VCUA achieves optimal timeliness during underloads, and tightly upper bounds inter and intratask completion times. Our simulation experiments confirm the algorithms effectiveness and superiority