Akitsugu Watanabe

VLEI code: an efficient labeling method for handling XML documents in an RDB

A number of XML labeling methods have been proposed to store XML documents in relational databases. However, they have a vulnerable point, in insertion operations. We propose the variable length endless insertable (VLEI) code and apply it to XML labeling to reduce the cost of insertion operations. Results of our experiments indicate that a combination of the VLEI code and Dewey order is effective for handling skewed insertions.

Adaptive overlapped declustering: a highly available data-placement method balancing access load and space utilization

This paper proposes a new data-placement method named adaptive overlapped declustering, which can be applied to a parallel storage system using a value range partitioning-based distributed directory and primary-backup data replication, to improve the space utilization by balancing their access loads. The proposed method reduces data skews generated by data migration for balancing access load. While some data-placement methods capable of balancing access load or reducing data skew have been proposed, both requirements satisfied simultaneously. The proposed method also improves the reliability and availability of the system because it reduces recovery time for damaged backups after a disk failure. The method achieves this acceleration by reducing a large amount of network communications and disk I/O. Mathematical analysis shows the efficiency of space utilization under skewed access workloads. Queuing simulations demonstrated that the proposed method halves backup restoration time, compared with the traditional chained declustering method.

On sequential diagnosis of multiprocessor systems

This paper considers the problem of sequential fault diagnosis for various multiprocessor systems. We propose a simple sequential diagnosis algorithm and show that the degree of sequential diagnosability of any system with N processors is at least Ω(√N). We also show upper bounds for various networks. These are the first nontrivial upper bounds for the degree of sequential diagnosability, to the best of our knowledge. Our upper bounds are proved in a unified manner, which is based on the very definition of sequential diagnosability. We show that a d-dimensional grid and torus with N vertices are sequentially O(Nd/(d+1))-diagnosable, and an N-vertex k-ary tree is O(√kN)- diagnosable. Moreover, we prove that the degree of sequential diagnosability of an N-vertex hypercube is at least Ω(N/√log N) and at most O(N log log N/√log N), and those of an N-vertex CCC, shuffle-exchange graph, and de Bruijn graph are Θ(N/log N).

A high-availability software update method for distributed storage systems

SUMMARY In this paper, we propose a nonstop system upgrademethod without significant performance degradation fordata management software. To reduce disk accesses andnetwork traffic, we construct logical nodes inside a physicalnode and migrate data between the symbiotic logical nodes.This logical migration is assisted by storage managementfunctions that hide data location and migration from users.We also show the effectiveness of our method using experi-mental results on the system based on the AutonomousDisks we have proposed as a highly available storage sys-tem technology.

Directory-traverse-cost-based skew handling for parallel data access

Techniques for load balancing by adjusting location of data are widely studied, because they are quite influential in improving the data-access performance and the scalability of a parallel system. To make load balancing effective, distributed directory structures, methods for evaluating loads, and distributed control mechanism for handling skews are important. In this paper, we propose a distributed algorithm to evaluate loads precisely. It counts the loads of intermediate index nodes of a distributed directory, which were ignored in previous publications. We also propose a linear-order distributed skew-handling algorithm adopting the proposed load evaluation algorithm. We validate the effect of our load evaluation algorithm through simulations.

Compound treatment of chained declustered replicas using a parallel btree for high scalability and availability

Scalability and availability are key features of parallel database systems. To realize scalability, many dynamic load-balancing methods with data placement and parallel index structures on shared-nothing parallel infrastructure have been proposed. Data migration with range-partitioned placement using a parallel Btree is one solution. The combination of range partitioning and chained declustered replicas provides high availability while preserving scalability. However, independent treatment of the primary and backup data in each node results in long failover times. We propose a novel method for compound treatment of chained declustered replicas using a parallel Btree, called the Fat-Btree. In the proposed method, the single Fat-Btree provides access paths to both primary and backup data in all processor elements, which greatly reduces failover time. Moreover, it enables dynamic load balancing without physical data migration, and improves memory space utilization for processing the index. Experiments using PostgreSQL on a 160-node PC cluster demonstrate the effect.

An Efficient Access Forwarding Method Based On Caches On Storage Nodes

In this paper, we discuss access forwarding schemes for the replication that achieve balanced access load distribution for data replicas on multiple storage nodes. In parallel storage systems, it is significant to handle skew of accessrequest distribution. Although replication is commonly used to solve such problems, it decreases hit ratios of cache memories on each storage nodes. We first obtain a result that it uses cache memories efficiently to forward accesses that target at data in less demand with a simple experiment. We also propose a method that uses hit ratios of each cache to recognize the popularity of data with low costs. Results of experiments with the method show that it helps the replication use the limited cache space efficiently.

The Versioning System Balancing Data Amount and Access Frequency on Distributed Storage System

In this paper, a method of balancing both access frequency and data amount for a distributed parallel storage system under version management is discussed. We assume the version management method keeps the latest version of file and a number of differential information sets to access previous versions. Since the access frequency for an aged version is tend to be lower than that for the latest version, we control the access frequency distribution by the placement of the latest versions, while the data amount distribution is managed by the placement of the differential information whose size is enough small to adjust the subtle difference of data amount. We propose a distributed directory structure and data placement algorithm, and evaluate its effect on the access frequency and data amount distribution.