Cluster-Based Servers             Event Type Start Time End Time Rm # Chair     Paper 10:30AM 11:00AM 36-37 Ruth Aydt (NCSA)   Title: An Efficient Data Location Protocol for Self-organizing Storage Clusters   Speakers/Presenter: Hong Tang (University of California, Santa Barbara), Tao Yang (University of California, Santa Barbara)   Paper 11:00AM 11:30AM 36-37 Ruth Aydt (NCSA)   Title: Handling Heterogeneity in Shared-Disk File Systems   Speakers/Presenter: Changxun Wu (Johns Hopkins University), Randal Burns (Johns Hopkins University)   Paper 11:30AM 12:00PM 36-37 Ruth Aydt (NCSA)   Title: Quantifying and Improving the Availability of High-Performance Cluster-Based Internet Services   Speakers/Presenter: Kiran Nagaraja (Computer Science, Rutgers University), Neeraj Krishnan (Computer Science, Rutgers University), Ricardo Bianchini (Computer Science, Rutgers University), Richard P. Martin (Computer Science, Rutgers University), Thu D. Nguyen (Computer Science, Rutgers University)

Session: Cluster-Based Servers   Title: An Efficient Data Location Protocol for Self-organizing Storage Clusters   Chair: Ruth Aydt (NCSA)   Time: Tuesday, November 18, 10:30AM - 11:00AM   Rm #: 36-37   Speaker(s)/Author(s):     Hong Tang (University of California, Santa Barbara), Tao Yang (University of California, Santa Barbara)       Description:   Component additions and failures are common for large-scale storage clusters in production environments. To improve availability and manageability, we investigate and compare data location schemes for a large self-organizing storage cluster that can quickly adapt to the additions or departures of storage nodes. We further present an efficient location scheme that differentiates between small and large file blocks for reduced management overhead compared to uniform strategies. In our protocol, small blocks, which are typically in large quantities, are placed through consistent hashing. Large blocks, much fewer in practice, are placed through a usage-based policy, and their locations are tracked by Bloom filters. The proposed scheme results in improved space utilization even with non-uniform cluster nodes. To achieve high scalability and fault resilience, this protocol is fully distributed, relies only on soft states, and supports data replication. We demonstrate the effectiveness and efficiency of this protocol through trace-driven simulation. This paper has been nominated for the Best Student Paper of SC2003 award.   Link: Download PDF

Session: Cluster-Based Servers   Title: Handling Heterogeneity in Shared-Disk File Systems   Chair: Ruth Aydt (NCSA)   Time: Tuesday, November 18, 11:00AM - 11:30AM   Rm #: 36-37   Speaker(s)/Author(s):     Changxun Wu (Johns Hopkins University), Randal Burns (Johns Hopkins University)       Description:   We develop and evaluate a system for load management in shared-disk file systems built on clusters of heterogeneous computers. The system generalizes load balancing and server provisioning. It balances file metadata workload by moving file sets among cluster server nodes. It also responds to changing server resources that arise from failure and recovery and dynamically adding or removing servers. The system is adaptive and self-managing. It operates without any a-priori knowledge of workload properties or the capabilities of the servers. Rather, it continuously tunes load placement using a technique called adaptive, non-uniform (ANU) randomization. ANU randomization realizes the scalability and metadata reduction benefits of hash-based, randomized placement techniques. It also avoids hashing's drawbacks: load skew, inability to cope with heterogeneity, and lack of tunability. Simulation results show that our load-management algorithm performs comparably to a prescient algorithm.   Link: Download PDF

Session: Cluster-Based Servers   Title: Quantifying and Improving the Availability of High-Performance Cluster-Based Internet Services   Chair: Ruth Aydt (NCSA)   Time: Tuesday, November 18, 11:30AM - 12:00PM   Rm #: 36-37   Speaker(s)/Author(s):     Kiran Nagaraja (Computer Science, Rutgers University), Neeraj Krishnan (Computer Science, Rutgers University), Ricardo Bianchini (Computer Science, Rutgers University), Richard P. Martin (Computer Science, Rutgers University), Thu D. Nguyen (Computer Science, Rutgers University)       Description:   Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. We apply a quantification methodology to show, however, that following a cooperative strategy without additional compensating mechanisms results in a substantial availability loss. Specifically, we show that a sophisticated cooperative cluster-based web server gains a factor of 3 in performance but increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.   Link: Download PDF