File Virtualization with DirectNFS Anupam Bhide, Anu Engineer, Anshuman Kanetkar, Aditya Kini {an... more File Virtualization with DirectNFS Anupam Bhide, Anu Engineer, Anshuman Kanetkar, Aditya Kini {anupam, anu, anshuman, aditya}@calsoftinc.com ... The Introduction of a simultaneous data access path can improve file serving performance through parallel and direct transfer of ...
SIGLEAvailable from British Library Document Supply Centre-DSC:4335.26205(2001-19) / BLDSC - Brit... more SIGLEAvailable from British Library Document Supply Centre-DSC:4335.26205(2001-19) / BLDSC - British Library Document Supply CentreGBUnited Kingdo
Practical experience indicates that the definition of real-world workflow applications is a compl... more Practical experience indicates that the definition of real-world workflow applications is a complex and error-prone process. Existing workflow management systems provide the means, in the best case, for very primitive syntactic verification, which is not enough to guarantee the overall correctness and robustness of workflow applications. The report introduces a method for formal verification of system behaviour which, in C3DS, is defined as workflow schemas. Workflows are modelled by means of an automata-based method, which facilitates exhaustive compositional reachability analysis. The workflow behaviour is checked against safety properties, which can be either generic (applicable to all workflow schemas) or domain specific (applicable to a given schema). The analysis is performed in an automated way using the LTSA toolkit, which is accessible by designers who are not experts in formal methods.
24th International Conference on Distributed Computing Systems, 2004. Proceedings., 2004
Data replication is used extensively in wide-area dis-tributed systems to achieve low data-access... more Data replication is used extensively in wide-area dis-tributed systems to achieve low data-access latency. A large number of heuristics have been proposed to perform replica placement. Practical experience indicates that the choice of heuristic makes a big difference in terms of the ...
Proceedings of the 10th workshop on ACM SIGOPS European workshop: beyond the PC - EW10, 2002
Pangaea is a planetary-scale file system designed for large, multi-national corporations or group... more Pangaea is a planetary-scale file system designed for large, multi-national corporations or groups of collaborating users spread over the world. Its goal is to handle people's daily stor-age needs, e. g., document sharing, software development, and data crunching, that can be write intensive. Pangaea uses per-vasive replication to achieve low access latency and high avail-ability. It creates replicas dynamically whenever
Proceedings. 14th Symposium on Reliable Distributed Systems, 1995
T h e paper addresses the problem of providing highly available services in distributed systems. ... more T h e paper addresses the problem of providing highly available services in distributed systems. In particular, we examine the situation where a service m a y be used by a large continuously changing set of clients. T h e requirements for providing services in this environment are analysed and a n architecture and partial implementation for a replicated server group meeting a range of client requirements is presented. T h e archi-
Proceedings of the 5th symposium on Operating systems design and implementation - OSDI '02, 2002
Pangaea is a wide-area file system that supports data shar-ing among a community of widely distri... more Pangaea is a wide-area file system that supports data shar-ing among a community of widely distributed users. It is built on a symmetrically decentralized infrastructure that consists of commodity computers provided by the end users. Computers act autonomously to serve data to their ...
Twelfth IEEE International Workshop on Quality of Service, 2004. IWQOS 2004., 2004
Ensuring performance isolation and differentiation among workloads that share a storage infrastru... more Ensuring performance isolation and differentiation among workloads that share a storage infrastructure is a basic requirement in consolidated data centers. Existing management tools rely on resource provisioning to meet performance goals; they depend on detailed knowledge of the system characteristics and the workloads. Provisioning is inherently slow to react to system and workload dynamics, and in the general case, it is impossible to provision for the worst case.
2005 International Conference on Control and Automation, 2005
Shared computing services must control resource usage to meet contractual performance goals for h... more Shared computing services must control resource usage to meet contractual performance goals for hosted customers. They must ensure performance isolation among the workloads of different customers and enforce prioritization when the service is overloaded. Existing solutions are domain-specific and require modifications to the service. We propose a generic, non-intrusive approach that uses a fair scheduler to intercept incoming requests and enforce proportional sharing of the service resources among workloads. The relation between workload shares and obtained performance varies over time depending on system dynamics. We design an adaptive optimal MIMO controller that dynamically sets the workload shares based on the observed performance. Experimental results from an NFS server and a 3-tier e-commerce site demonstrate that the controller achieves effective performance differentiation, even when the system state or the performance goals change significantly.
Industry analysts see no limits to the world's expanding appetite for data-storage services. ... more Industry analysts see no limits to the world's expanding appetite for data-storage services. Emerging networking technologies allow incremental scaling of bandwidth and capacity of storage, which is attached to the network. A key challenge is to devise the software that provides transparent shared access to decentralized storage resources. Existing network file systems will not meet the scalability requirements of future
File Virtualization with DirectNFS Anupam Bhide, Anu Engineer, Anshuman Kanetkar, Aditya Kini {an... more File Virtualization with DirectNFS Anupam Bhide, Anu Engineer, Anshuman Kanetkar, Aditya Kini {anupam, anu, anshuman, aditya}@calsoftinc.com ... The Introduction of a simultaneous data access path can improve file serving performance through parallel and direct transfer of ...
SIGLEAvailable from British Library Document Supply Centre-DSC:4335.26205(2001-19) / BLDSC - Brit... more SIGLEAvailable from British Library Document Supply Centre-DSC:4335.26205(2001-19) / BLDSC - British Library Document Supply CentreGBUnited Kingdo
Practical experience indicates that the definition of real-world workflow applications is a compl... more Practical experience indicates that the definition of real-world workflow applications is a complex and error-prone process. Existing workflow management systems provide the means, in the best case, for very primitive syntactic verification, which is not enough to guarantee the overall correctness and robustness of workflow applications. The report introduces a method for formal verification of system behaviour which, in C3DS, is defined as workflow schemas. Workflows are modelled by means of an automata-based method, which facilitates exhaustive compositional reachability analysis. The workflow behaviour is checked against safety properties, which can be either generic (applicable to all workflow schemas) or domain specific (applicable to a given schema). The analysis is performed in an automated way using the LTSA toolkit, which is accessible by designers who are not experts in formal methods.
24th International Conference on Distributed Computing Systems, 2004. Proceedings., 2004
Data replication is used extensively in wide-area dis-tributed systems to achieve low data-access... more Data replication is used extensively in wide-area dis-tributed systems to achieve low data-access latency. A large number of heuristics have been proposed to perform replica placement. Practical experience indicates that the choice of heuristic makes a big difference in terms of the ...
Proceedings of the 10th workshop on ACM SIGOPS European workshop: beyond the PC - EW10, 2002
Pangaea is a planetary-scale file system designed for large, multi-national corporations or group... more Pangaea is a planetary-scale file system designed for large, multi-national corporations or groups of collaborating users spread over the world. Its goal is to handle people's daily stor-age needs, e. g., document sharing, software development, and data crunching, that can be write intensive. Pangaea uses per-vasive replication to achieve low access latency and high avail-ability. It creates replicas dynamically whenever
Proceedings. 14th Symposium on Reliable Distributed Systems, 1995
T h e paper addresses the problem of providing highly available services in distributed systems. ... more T h e paper addresses the problem of providing highly available services in distributed systems. In particular, we examine the situation where a service m a y be used by a large continuously changing set of clients. T h e requirements for providing services in this environment are analysed and a n architecture and partial implementation for a replicated server group meeting a range of client requirements is presented. T h e archi-
Proceedings of the 5th symposium on Operating systems design and implementation - OSDI '02, 2002
Pangaea is a wide-area file system that supports data shar-ing among a community of widely distri... more Pangaea is a wide-area file system that supports data shar-ing among a community of widely distributed users. It is built on a symmetrically decentralized infrastructure that consists of commodity computers provided by the end users. Computers act autonomously to serve data to their ...
Twelfth IEEE International Workshop on Quality of Service, 2004. IWQOS 2004., 2004
Ensuring performance isolation and differentiation among workloads that share a storage infrastru... more Ensuring performance isolation and differentiation among workloads that share a storage infrastructure is a basic requirement in consolidated data centers. Existing management tools rely on resource provisioning to meet performance goals; they depend on detailed knowledge of the system characteristics and the workloads. Provisioning is inherently slow to react to system and workload dynamics, and in the general case, it is impossible to provision for the worst case.
2005 International Conference on Control and Automation, 2005
Shared computing services must control resource usage to meet contractual performance goals for h... more Shared computing services must control resource usage to meet contractual performance goals for hosted customers. They must ensure performance isolation among the workloads of different customers and enforce prioritization when the service is overloaded. Existing solutions are domain-specific and require modifications to the service. We propose a generic, non-intrusive approach that uses a fair scheduler to intercept incoming requests and enforce proportional sharing of the service resources among workloads. The relation between workload shares and obtained performance varies over time depending on system dynamics. We design an adaptive optimal MIMO controller that dynamically sets the workload shares based on the observed performance. Experimental results from an NFS server and a 3-tier e-commerce site demonstrate that the controller achieves effective performance differentiation, even when the system state or the performance goals change significantly.
Industry analysts see no limits to the world's expanding appetite for data-storage services. ... more Industry analysts see no limits to the world's expanding appetite for data-storage services. Emerging networking technologies allow incremental scaling of bandwidth and capacity of storage, which is attached to the network. A key challenge is to devise the software that provides transparent shared access to decentralized storage resources. Existing network file systems will not meet the scalability requirements of future
Uploads
Papers by C. Karamanolis