Riska, Alma - Seagate Research Center

• E#cient fitting of longtailed data sets into PH distributions #
• It's not fair -evaluating efficient disk scheduling Alma Riska Erik Riedel
• Disk Drive Workload Captured in Logs Collected During the Field Return Incoming Test
• International Conference on Dependable Systems & Networks: Anchorage, Alaska, June 24-27 2008 Enhancing Data Availability in Disk Drives through Background Activities*
• Evaluating Block-level Optimization through the IO Path Seagate Research
• Long-Range Dependence at the Disk Drive Level Alma Riska Erik Riedel
• Evaluating the Performability of Systems with Background Jobs , Alma Riska2
• Storage Performance Virtualization via Throughput and Latency Control Jianyong Zhang Alma Riska Anand Sivasubramaniam Qian Wang Erik Riedel
• Adaptive disk scheduling for overload management Seagate Research
• Managing overload via adaptive scheduling Seagate Research
• Bridging ETAQA and Ramaswami's formula for the solution of M/G/1-type processes
• Efficient fitting of long-tailed data sets into PH distributions
• Efficient fitting of long-tailed data sets into hyperexponential distributions Alma Riska Vesselin Diev Evgenia Smirni
• Susceptibility of Commodity Systems and Software to Memory Soft Errors
• Efficient fitting of longtailed data sets into hyperexponential distributions Alma Riska Vesselin Diev Evgenia Smirni
• Characterization of the E-commerce Storage Subsystem Workload The Ohio State University
• Exact Aggregate Solutions for M/G/1type Markov Processes #
• Exact analysis of a class of GI/G/1-type performability Alma Riska Evgenia Smirni Gianfranco Ciardo
• AdaptLoad: e ective balancing in clustered web servers under transient load conditions
• [28] P. M. Snyder and W. J. Stewart. Explicit and iterative numerical approaches to solving queueing models. Operations Research, 33(1):183{202, Jan-Feb. 1985.
• Idle Read After Write -IRAW Seagate Research
• MAMSolver: A matrix analytic methods tool 1 Alma Riska Evgenia Smirni
• Workload Propagation Overload in Bursty Servers Computer Science Dept.
• Analytic Modeling of Load Balancing Policies for Tasks with Heavytailed Distributions \Lambda
• Evaluation of Disk-level Workloads at Different Time-scales College of William and Mary
• Load Unbalancing to Improve Performance under Autocorrelated Traffic Qi Zhang Ningfang Mi
• Exact Aggregate Solutions for M/G/1-type Markov Alma Riska Evgenia Smirni
• Bottlenecks and their Performance Implications in E-Commerce Systems
• Performance-Guided Load (Un)Balancing under Autocorrelated Flows
• ARTICLE IN PRESS Performance Evaluation ( )
• JVM Susceptibility to Memory Errors Deqing Chen
• Disk Drive Level Workload Characterization Seagate Research
• WorkloadAware Load Balancing for Clustered Web Qi Zhang Alma Riska Wei Sun Evgenia Smirni Gianfranco Ciardo
• M/G/1-type Markov Processes: A tutorial Alma Riska and Evgenia Smirni
• Etaqa-MG1: An ecient technique for the analysis of a class of M/G/1-type processes by aggregation
• Restrained Utilization of Idleness for Transparent Scheduling of Background Tasks
• An aggregationbased solution method for M/G/1type processes
• Load Balancing for Performance Differentiation in Dual-Priority Clustered Ningfang Mi Qi Zhang
• Storage Performance Virtualization via Throughput and Latency Control
• M/G/1type Markov Processes: A tutorial # Alma Riska and Evgenia Smirni
• MATRIX-ANALYTIC ANALYSIS OF A MAP/PH/1 QUEUE FITTED TO WEB SERVER DATA
• An Interposed 2-Level I/O Scheduling Framework for Performance Virtualization
• Efficient Management of Idleness in Storage Ningfang Mi1
• Efficient Management of Idleness in Systems Ningfang Mi
• Bridging ETAQA and Ramaswami's formula for the solution of M/G/1type processes
• Workload-Aware Load Balancing for Clustered Web Servers