Shailabh Nagar

A closer look at coscheduling approaches for a network of workstations

Efficient scheduling of processes on processors of a Network of Workstations (NOW) is essential for good system performance. However, the design of such schedulers is challenging because of the complex interaction between several system and workload parameters. Coscheduling, though desirable, is impractical for such a loosely coupled environment. Two operations, waiting for a message and arrival of a message, can be used to take remedial actions that can guide the behavior of the system towards coscheduling using local information. We present a taxonomy of three possibilities for each of these two operations, leading to a design space of scheduling mechanisms. This paper presents an extensive implementation and evaluation exercise in studying these mechanisms. Adhering to the philosophy that scheduling and communication are intertwined and should be studied in conjunction, a complete communication substrate for UltraSPARC workstations, connected by Myrinet and running Solaris 2.5.1, has been developed. This platform provides the entire Message Passing Interface (MPI) to readily run off-the-shelf MPI applications by employing protected low-latency user-level messaging. Several applications can concurrently use this interface. This platform has been used to design, implement, and uniformly evaluate nine scheduling strategies with a mixture of concurrent real applications with varying communication intensities. This includes four new schemes (Periodic Boost, Periodic Boost with Spin Block, Spin Yield, Periodic Boost with Spin Yield) that are presented in this paper. In addition to evaluatThis research is supported in part by an NSF Career Award MIPS-9701475, NSF grant MIP-9634197 and an NSF equipment grant CDA-9617315. ing the pros and cons of each mechanism in terms of throughput, response time, CPU utilization and fairness, it is shown that Periodic Boost is a promising approach for scheduling processes on a NOW.

Synthesizing Representative I/O Workloads for TPC-H

Synthesizing I/O requests that can accurately capture workload behavior is extremely valuable for the design, implementation and optimization of disk subsystems. This paper presents a synthetic workload generator for TPC-H, an important decision-support commercial workload, by completely characterizing the arrival and access patterns of its queries. We present a novel approach for parameterizing the behavior of inter-mingling streams of sequential requests, and exploit correlations between multiple attributes of these requests, to generate disk block-level traces that are shown to accurately mimic the behavior of a real trace in terms of response time characteristics for each TPC-H query.

Alternatives to Coscheduling a Network of Workstations

Efficient scheduling of processes on processors of a Network of Workstations (NOW) is essential for good system performance. However, the design of such schedulers is challenging because of the complex interaction between several system and workload parameters. Coscheduling, though desirable, is impractical for such a loosely coupled environment. Two operations, waiting for a message and arrival of a message, can be used to take remedial actions that can guide the behavior of the system toward coscheduling using local information. We present a taxonomy of three possibilities for each of these two operations, leading to a design space of 3×3 scheduling mechanisms. This paper presents an extensive implementation and evaluation exercise in studying these mechanisms. Adhering to the philosophy that scheduling and communication are intertwined and should be studied in conjunction, a complete communication substrate for UltraSPARC workstations, connected by Myrinet and running Solaris 2.5.1, has been developed. This platform provides the entire Message Passing Interface (MPI) to readily run off-the-shelf MPI applications by employing protected low-latency user-level messaging. Several applications can concurrently use this interface. This platform has been used to design, implement, and uniformly evaluate nine scheduling strategies with a mixture of concurrent real applications with varying communication intensities. This includes five new schemes (Periodic Boost, Periodic Boost with Spin Block, Spin Yield, Periodic Boost with Spin Yield, Dynamic Coscheduling with Spin Yield) that are presented in this paper. In addition to our evaluations of the pros and cons of each mechanism in terms of throughput, response time, CPU utilization, and fairness, it is shown that Periodic Boost is a promising approach for scheduling processes on a NOW.

Issues in designing and implementing a scalable virtual interface architecture

The Virtual Interface Architecture brings the benefits of low latency User-level Networking to a cluster environment. With an increasing number of communication channels created and used, the performance degradation of existing channels becomes an important consideration. This paper addresses the crucial issue of VIA scalability by examining the hardware and software options at the network interface and the host. Different design alternatives are investigated indepth using a detailed simulation model to get revealing insights from the scalability perspective.

Optimal control point selection for continuous business process compliance monitoring

Service delivery organizations fulfill their business obligations by defining and implementing business processes. Such processes also need to adhere to several regulations such as security, confidentiality and data integrity. These regulations are typically defined as policies, each of which contains a list of clauses. Organizations typically conduct periodic audits of executed process instances to determine how well they have adhered to the stated policies. Since such auditing takes place after the processes have been implemented, their efficacy as a tool for preventing non-compliances against policies is doubtful. In this paper, we analyze the problem of continuous compliance monitoring at run time. This involves selecting a subset of policy clauses at process tasks, called control points, at which compliance can be checked. Selecting control points involves a tradeoff. Selecting too few control points would raise the risk of increased non-compliance; however, selecting too many may not be cost-effective. In this paper, we represent this as an optimization problem, and discuss it from the viewpoint of two key optimization criteria. We also demonstrate our optimization techniques on a simplified version of a real-life example drawn from our industry experience.

Process Trace Identification from Unstructured Execution Logs

Many real world business processes are executed without explicit orchestration and hence do not generate structured execution logs. This is particularly true for the class of business processes which are executed in service delivery centers in emerging markets where rapid changes in processes and in the people executing the processes are common. In such environments, the process execution logs are usually natural language descriptions of actions performed and hence are noisy. Despite the lack of structured logs, it is crucial to know the trace of activities as they happen on the ground. Without such a visibility into the ground activities, regulatory compliance audit, process optimization, and best practices standardization are severely disabled. Process monitoring on top of unstructured execution logs has been a relatively unexplored research area. This paper proposes an approach for process trace identification from unstructured logs that applies state-of-the-art text mining techniques. It applies this approach on logs of a real-world business process used in a service delivery center and shows that individual process activities are correctly identified 90% of the time. Also, 65% of the activity traces were identified with zero errors and an additional 24% with a single error. This approach is generic and applicable to a wide array of business processes.

Indra: an integrated quantitative system for compliance management for IT service delivery

The explosive growth of business process implementations in various industries has brought into sharp focus the need for process compliance with regulatory policies. This has raised the need for business process compliance solutions requiring an automated and quantitative approach. Quantification of compliance enables an organization to accurately determine its compliance posture and take steps to improve process noncompliances in the future. To that end, in this paper, we propose Indra, our system for integrated compliance management. Indra takes a holistic approach toward compliance, focusing on a compliance life cycle comprising process modeling for maximal compliance at minimal cost, measuring noncompliance at runtime, analyzing the results of the measurement, and suggesting corrective actions to continuously improve process compliance in the future. The scope of this paper covers the analytic models and formulations for compliance maximization, along with a demonstration on a simplified version of a real-life example drawn from the IBM IT (information technology) service delivery units. We also describe ongoing piloting of our analytic models on real audit data from the IBM India Business Controls department. To the best of our knowledge, Indra is the first of its kind in providing integrated and quantitative compliance management.

Enabling automatic workload management in Linux

Complexity reduction in workload management is driving the development of goal-oriented workload managers (WLMs). Simultaneously, server consolidation of workloads with dynamically changing resource demands calls for these WLMs to be increasingly efficient in managing resources. We propose the class-based kernel resource management (CKRM) framework, implemented in Linux, for operating systems to these requirements.

Implementing Protected Multi-User Communication for Myrinet

A Network of Workstations (NOW) is emerging as a cost-effective solution to high performance computing. However, we need to lower the cost of communicating between the workstations to make this platform viable. With the advent of high-performance networks such as ATM and Myrinet, the physical network is no longer the communication bottleneck. Rather, the major overhead can now be attributed to software. This overhead is a direct result of the cost that a message incurs as it travels through different protection domains. We can alleviate this problem by allowing protected user-level access directly to the network, thereby eliminating the kernel from the critical path. This paper presents a description of the design, implementation and performance of a protected user-level messaging system over Myrinet, called MU-Net, that can handle multiple application processes concurrently. MU-Net has been implemented on the SUN Solaris 2.5 operating system.

pSNOW: a tool to evaluate architectural issues for NOW environments

Performance evaluation plays a crucial role in the design of any system. Evaluation tools should clearly identify, isolate and quantify the bottlenecks in the execution to help restructure the application for better performance, as well as suggest enhancements to the existing design. While there has been significant progress recently in novel network interface designs and system software solutions to lower the communication overheads on emerging high performance Network of Workstations environments, performance evaluation tools for these environments have not kept pace with this progress. In this research, we present an execution-driven simulation tool called pSNOW that provides us a unified framework to model different system software and architectural designs, and evaluate these designs using real applications. Using this tool, we model three network interfaces and three communication software substrates, and evaluate their relative merits and demerits.