Search

Continuous queries over data streams have been used in many applications. They incrementally produce their query results according to data arrival and time progress. However, existing continuous query languages strongly assume that all information sources to be accessed by queries are explicitly specified and permanently fixed. There are more sophisticated requirements such as changing information sources and rewriting filter conditions during query execution. This paper proposes a dynamic query modification scheme for continuous queries. Our continuous query language treats not only stream data from information sources, but also metadata of the information sources and queries themselves. Internally, a continuous query is represented as a dataflow, which is an RDF graph of operators. Thus, modifying a continuous query can be regarded as rewriting the corresponding RDF graph. Searching information sources which satisfy the conditions can be also regarded as searching RDF graphs descripting metadata of information sources. Our scheme achieves both continuous searching information sources and rewriting dataflows.

SUMMARY The numbers of files in file systems have increased dramatically in recent years. Office workers spend much time and effort searching for the documents required for their jobs. To reduce these costs, we propose a new method for recommending files and operations on them. Existing technologies for recommendation, such as collaborative filtering, suffer from two problems. First, they can only work with documents that have been accessed in the past, so that they cannot recommend when only newly generated documents are inputted. Second, they cannot easily handle sequences involving similar or differently ordered elements because of the strict matching used in the access sequences. To solve these problems, such minor variations should be ignored. In our proposed method, we introduce the concepts of abstract files as groups of similar files used for a similar purpose, abstract tasks as groups of similar tasks, and frequent abstract workflows grouped from similar workflows, which are sequences of

It is useful to employ electronic medical records to improve medical studies. Based on their experience, medical workers conventionally prepare clinical pathways as guidelines for the typical flow for the medical treatment of each disease. In this study, we propose an approach for verifying existing clinical pathways and recommend variants or new pathways by analyzing historical records. We propose a method based on the application of sequential pattern mining to record logs with handling time intervals between treatments. We also focus on the efficacy of medicines instead of their names because various medicines have the same efficacy and they change dynamically. We evaluated the proposed method using actual logs and the results demonstrated that the proposed method is effective.

We report a method for the coacervation micro-encapsulation of several forms of CaCO3 microparticles with the fluoropolymer poly(heptadecafluorodecyl acrylate) (poly (HDFDA)) by pressure-induced phase separation of a supercritical CO2 solution. A suspension of CaCO3 in CO2 and dissolved poly(HDFDA) were mixed in supercritical CO2. After the system pressure was slowly decreased to atmospheric pressure, the microcapsules were obtained. Coacervation was achieved by the precipitation of poly(HDFDA) during the decrease in the pressure of CO2; the solubility of poly(HDFDA) in CO2 decreased with the pressure. The structure and morphology of the microparticles were investigated by using a scanning electron microscope (SEM) and an electron probe microanalyzer (EPMA) equipped with a wavelength dispersive X-ray spectroscope (WDX).

Hadoop has been widely used in various clusters to build scalable and high performance distributed file systems. However, Hadoop distributed file system (HDFS) is designed for large file management. In case of small files applications, those metadata requests will flood the network and consume most of the memory in Namenode thus sharply hinders its performance. Therefore, many web applications do not benefit from clusters with centered metanode, like Hadoop. In this paper, we compare our Fat-Btree based data access method, which excludes center node in clusters, with Hadoop. We show their different performance in different file I/O applications.

Files in computers are increasing in number, so we require file management tools to find target files and to classify large groups of files. Our research group has been developing a system that provides virtual directories made up of related files. Many methods to extract inter-file relationships are available, such as word frequency, access co-occurrence, and so on. In practice, users need to select and combine multiple types of inter-file relationships to form groups of files they want. For this purpose, we propose FileSearchCube, a tool for file group analysis that introduces the data-cube concept. FileSearchCube provides cube operations for a set of files, and helps users to find relevant file groups.

Interest in the Web services (WS) composition (WSC) paradigm is increasing tremendously. A real shift in distributed computing history is expected to occur when the dream of implementing Service-Oriented Architecture (SOA) is realized. However, there is a long way to go to achieve such an ambitious goal. In this paper, we support the idea that, when challenging the WSC issue, the earlier that the inevitability of failures is recognized and proper failure-handling mechanisms are defined, from the very early stage of the composite WS (CWS) specification, the greater are the chances of achieving a significant gain in dependability. To formalize this vision, we present the FENECIA (Failure Endurable Nested-transaction based Execution of Composite Web services with Incorporated state Analysis) framework. Our framework approaches the WSC issue from different points of view to guarantee a high level of dependability. In particular, it aims at being simultaneously a failure-handling-devoted CWS specification, execution, and quality of service (QoS) assessment approach. In the first section of our framework, we focus on answering the need for a specification model tailored for the WS architecture. To this end, we introduce WS-SAGAS, a new transaction model. WS-SAGAS introduces key concepts that are not part of the WS architecture pillars, namely, arbitrary nesting, state, vitality degree, and compensation, to specify failure-endurable CWS as a hierarchy of recursively nested transactions. In addition, to define the CWS execution semantics, without suffering from the hindrance of an XML-based notation, we describe a textual notation that describes a WSC in terms of definition rules, composability rules, and ordering rules, and we introduce graphical and formal notations. These rules provide the solid foundation needed to formulate the execution semantics of a CWS in terms of execution correctness verification dependencies. To ensure dependable execution of the CWS, we present in the second section of FENECIA our architecture THROWS, in which the execution control of the resulting CWS is distributed among engines, discovered dynamically, that communicate in a peer-to-peer fashion. A dependable execution is guaranteed in THROWS by keeping track of the execution progress of a CWS and by enforcing forward and backward recovery. We concentrate in the third section of our approach on showing how the failure consideration is trivial in acquiring more accurate CWS QoS estimations. We propose a model that assesses several QoS properties of CWS, which are specified as WS-SAGAS transactions and executed in THROWS. We validate our proposal and show its feasibility and broad applicability by describing an implemented prototype and a case study.

With the exponential increase in the amount of XML data on the Internet, information retrieval techniques on tree-structured XML documents such as keyword search become important. The search results for this retrieval technique are often represented by minimum connecting trees (MCTs) rooted at the lowest common ancestors (LCAs) of the nodes containing all the search keywords. Recently, effective methods such as the stack-based algorithm for generating the lowest grouped distance MCTs (GDMCTs), which derive a more compact representation of the query results, have been proposed. However, when the XML documents and the number of search keywords become large, these methods are still expensive. To achieve more efficient algorithms for extracting MCTs, especially lowest GDMCTs, we first consider two straightforward LCA detection methods: keyword B+trees with Dewey-order labels and superimposed code-based indexing methods. Then, we propose a method for efficiently detecting the LCAs, which combines the two straightforward indexing methods for LCA detection. We also present an effective solution for the false drop problem caused by the superimposed code. Finally, the proposed LCA detection methods are applied to generate the lowest GDMCTs. We conduct detailed experiments to evaluate the benefits of our proposed algorithms and show that the proposed combined method can completely solve the false drop problem and outperforms the stack-based algorithm in extracting the lowest GDMCTs.

In this paper, we propose an effective method for segmenting large XML documents into independent meaningful subtrees based on two syntactic segmentation rates: vertical segmentation rate and horizontal segmentation rate. In the proposed method, we use DO-VLEI code to calculate the required parameters for the subtree segmentation. We conduct experiments to observe the effectiveness of the proposed subtree segmentation method using real bibliography XML documents stored in RDBs. We apply our previously proposed subtree matching algorithm SLAX to match the segmented subtrees and evaluate how the matching threshold impacts the precision and recall of subtree matching. Besides, we also integrate the matched subtrees determined by SLAX by our previously proposed subtree integration algorithm. The experimental results indicate that the proposed subtree segmentation method is effective for segmenting XML documents into independent meaningful subtrees and our previously proposed subtree matching algorithm achieves reasonable matching precision and recall using the segmented subtrees.

We have previously proposed Autonomous Disks System as a dependable and high-performance networked parallel storage system. The system has function of load balancing by online horizontal data migration. However, the migration causes a reduction of system availability by insensitively utilizing of system resources for massive data transfers. Thus we have also proposed Replica-assisted Migration, which solve the problem by using replica data for data redundancy. In this paper we show the efficiency of this method with experimental results on simulation program, comparing to speed adjusting method which is widely used for service-aware data migration.

In this paper, we propose a new concurrency control protocol for parallel B-tree structures capable reducing the cost of structure-modification-operation (SMO) compared to the conventional protocols such as ARIES/IM and INC-OPT. We call this protocol the MARK-OPT protocol, since it marks the lowest SMO occurrence point during optimistic latch-coupling operations. The marking reduces middle phases for spreading an X latch and removes needless X latches. In addition, we propose three variations of the MARK-OPT, which focus on tree structure changes from other transactions. Moreover, the proposed protocols are deadlock-free and satisfy the physical consistency requirement for B-trees. These indicate that the proposed protocols are suitable as concurrency control protocols for B-tree structures. To compare the performance of the proposed protocols, the INC-OPT, and the ARIES/IM, we implement these protocols on an autonomous disk system adopting the Fat-Btree structure, a form of parallel B-tree structure. Experimental results in various environments indicate that the proposed protocols always improve system throughput, and 2P-REP-MARK-OPT is the most useful protocol in high update environment. Additionally, to mitigate access skew, data should be migrated between PEs. We also demonstrate that MARK-OPT improves the system throughput under the data migration and reduces the time for data migration to balance load distribution.

Parasporin-1 is a novel non-insecticidal inclusion protein from Bacillus thuringiensis that is cytotoxic to specific mammalian cells. In this study, we investigated the effects of parasporin-1 on toxin-sensitive cell lines to elucidate the cytotoxic mechanism of parasporin-1. Parasporin-1 is not a membrane pore-forming toxin as evidenced by measurements of lactate dehydrogenase release, propidium iodide penetration, and membrane potential in parasporin-1-treated cells. Parasporin-1 decreased the level of cellular protein and DNA synthesis in parasporin-1-sensitive HeLa cells. The earliest change observed in cells treated with this toxin was a rapid elevation of the intracellular free-Ca(2+) concentration; increases in the intracellular Ca(2+) levels were observed 1-3 min following parasporin-1 treatment. Using four different cell lines, we found that the degree of cellular sensitivity to parasporin-1 was positively correlated with the size of the increase in the intracellular Ca(2+) concentration. The toxin-induced elevation of the intracellular Ca(2+) concentration was markedly decreased in low-Ca(2+) buffer and was not observed in Ca(2+)-free buffer. Accordingly, the cytotoxicity of parasporin-1 decreased in the low-Ca(2+) buffer and was restored by the addition of Ca(2+) to the extracellular medium. Suramin, which inhibits trimeric G-protein signaling, suppressed both the Ca(2+) influx and the cytotoxicity of parasporin-1. In parasporin-1-treated HeLa cells, degradation of pro-caspase-3 and poly(ADP-ribose) polymerase was observed. Furthermore, synthetic caspase inhibitors blocked the cytotoxic activity of parasporin-1. These results indicate that parasporin-1 activates apoptotic signaling in these cells as a result of the increased Ca(2+) level and that the Ca(2+) influx is the first step in the pathway that underlies parasporin-1 toxicity.

Advanced data engineering applications require a largescale storage system that is both scalable and dependable. In such a system, an atomic commit protocol becomes imperative to ensure the consistency and atomicity of transactions. In this paper we present a new commit protocol, BA-1.5PC, which is well tailored to such distributed storage environments as autonomous disks [20] that use a primarybackup storage schema. The protocol achieves an efficient commit process while also guaranteeing a high dependability by combining several approaches: (1) a low-overhead log mechanism that eliminates blocking disk I/Os, (2) removing the voting phase from commit processing to gain a faster commit process, and (3) a primary-backup assisted recovery strategy to enhance dependability in the presence of possible failures, so that a master failure in the decision phase will not block prepared cohorts of a transaction. Experiments were carried out on a trial version of an autonomous disks system to verify its efficiency. The results indicate that this protocol significantly outperforms several well-known commit protocols in terms of transaction throughput.

The properties of contents stored in a computer system are very wide while the data volume treated in the system becomes very large. It is important to treat each stored object in different manners to reject its properties in the data management for the large amount of stored data. To satisfy the requirement, we propose a method for the autonomous management based on ECA rules stored in metadata of the contents. We study the feasibility of treating a large number of ECA rules corresponding to the number of stored objects. Because the cost for evaluating conditions in the rules becomes dominant to the system perfornzance when the number of objects increases, we divide the conditions into two types, previously evaluable conditions and nuttime evaluable conditions, and construct a discrimination network for the previously evaluable conditions of each event to reduce the cost for processing the rules. We implement the methods in the autonomous disk system, a high functional storage system we proposed, and evaluate the eficiency of them.

In this paper, we discuss the one-to-multiple matching problem in leaf-clustering based approximate XML join algorithms and propose a path-sequence based discrimination method to solve this problem. In our method, each path sequence from the top node to the matched leaf in the base and target subtree is extracted, and the most similar target subtree for the base one is determined by the pathsequence based subtree similarity degree. We conduct experiments to evaluate our method by using both real bibliography and bioinformatics XML documents. The experimental results show that our method can effectively decrease the occunence rate of one-to-multiple matching for both bibliography and bioinformatics XML data, and hence improve the precision of the leaf-clustering based approximate XML join algorithms.

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. © 2006 Wiley Periodicals, Inc. Syst CompJpn, 37(10): 35–46, 2006; Published online in Wiley Inter-Science (www.interscience.wiley.com). DOI 10.1002/scj.20503 Key words: network storage; high availability;ceaseless update; storage management. 1. Introduction Recently, although the complexity of software tomanage data on storage systems has increased becausesystems have become larger, the software also requirescontinued high availability.Large-scale storage systems are constructed from agreat number of magnetic hard disk drives (HDDs), ornodes, with consideration for both costs and performance.In such systems, the software to manage data providesseveral functions: storage virtualization to provide a unifiedstorage image, data redundancy control to protect dataunder disk failure, and so on.It is crucial that software is updated with no servicedowntime. The software may require updating after distri-bution for many reasons, such as bug fixes and scope creep.However, to avoid great losses of productivity, serviceopportunities, and so on, systems cannot be stopped.For the following discussion, we introduce two mod-els of relations between the location of the storage manage-ment and the data store

Recently, more and more data are published and exchanged by XML on the Internet. However, different XML data sources might contain the same data but have different structures. Therefore, it requires an efficient method to integrate such XML data sources so that more complete and useful information can be conveniently accessed and acquired by users. The tree edit distance is regarded as an effective metric for evaluating the structural similarity in XML documents. However, its computational cost is extremely expensive and the traditional wisdom in join algorithms cannot be applied easily. In this paper, we propose LAX (Leaf-clustering based Approximate XML join algorithm), in which the two XML document trees are clustered into subtrees representing independent items and the similarity between them is determined by calculating the similarity degree based on the leaf nodes of each pair of subtrees. We also propose an effective algorithm for clustering the XML document for LAX. We show that it is easily to apply the traditional wisdom in join algorithms to LAX and the join result contains complete information of the two documents. We then do experiments to compare LAX with the tree edit distance and evaluate its performance using both synthetic and real data sets. Our experimental results show that LAX is more efficient in performance and more effective for measuring the approximate similarity between XML documents than the tree edit distance.

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.

Over the last couple of years, the Web services compositions paradigm has been gathering a considerable momentum to grasp at the opportunity of becoming the most natural solution for autonomous and heterogeneous applications integration. In this paper, we advocate that we have yet to be concerned about the reliability and performance levels that the Web services composition will exhibit upon execution, which is still an ongoing research problem finding notable interest. Focusing on this issue, we propose to estimate the reliability and performance of Web services compositions. We concentrate on one important aspect that has received little attention so far: the consideration of the failures repercussions on the overall composition execution performance. Specifically, our targets are twofold. Firstly, we propose to enrich WS-SAGAS with a new set of advanced aggregation patterns so that it fits with the inherent business processes complexity and thereby it allows to specify, as Web services compositions, any business process no matter how it turns out to. Secondly, we propose to estimate the reliability and performance of each of these patterns, and this, in both, correct and faulty situations. Our enriched WS-SAGAS model ameliorates considerably the chances to acquire more plausible estimations of the reliability and performance because of its failure awareness. Moreover, analyzing these estimations helps to investigate the reasons that may lay behind these failures and indeed to contribute noticeably to acquire more reliable compositions execution with high performance level

The continued evolution of modern hardware has brought several new challenges to database management systems (DBMSs). Multicore CPUs are now mainstream, while the future lies in massively parallel computing performed on many-core processors. However, because they were developed originally for single-core processors, DBMSs cannot take full advantage of the parallel computing that uses so many cores. Several components in the traditional database engines become new bottlenecks on multicore platforms. In this paper, we analyze the bottlenecks in existing database engines on a modern multicore platform using the mixed workload of the TPC-W benchmark and describe strategies for higher scalability and throughput for existing DBMSs on multicore platforms. First, we show how to overcome the limitations of the database engine by introducing a partition-based multi-instance database system on a single multicore platform without any modification of existing DBMSs. Second, we analyze the possibility of further improving the performance by optimizing the cache performance of concurrent queries. Implemented by middleware, our proposed PM-DB can avoid the challenging work of modifying existing database engines. Performance evaluation using the TPC-W benchmark revealed that our proposal can achieve at most 2.5 times higher throughput than the existing engine of PostgreSQL.

Because of the development of information technologies and the pervasiveness of cloud services, Reducing power consumption in large scale storage systems becomes a very important issue. Recently, we have proposed a method to solve the problem called Replica Assisted Power Saving Disk Array (RAPoSDA) and verified its effectiveness. RAPoSDA utilizes a primary backup configuration on both cache memories and disk drives to ensure system reliability and it dynamically controls the timing and targeting of disk access based on individual disk rotation states. Until now, we have evaluated the effectiveness of RAPoSDA by using simulation program which we had developed. The evaluation shows that RAPoSDA achieves significantly reducing power consumption. However, we have not evaluated it effectiveness on the practical environment. In this paper, we introduce the prototype system of RAPoSDA to evaluate the power saving capability of RAPoSDA.

When the number of disks rises in disk array systems that contain multiple disk drives, system performance is limited by a bottleneck at a centralized controller and/or at a communication path that uses a bus. We evaluate, through simulation, a scalable architecture called DR-net, in which the controller functions are distributed to all disk drives and each disk has autonomy in processing its tasks. DR-net provided high throughput in proportion to the number of disks. In a conventional system, the influence of the bus setup time and concentration of the parity calculation load caused the throughput to saturate. We also show that DR-net can take advantage of disk autonomy when reconstructing data stored in failed disks. Our results indicate that the distribution of functions and the autonomy of disk drives enable better scalability and more effective utilization of system resources than with a hierarchical system.

Recently, it is needed to have high performance and use capacity efficiently at the same time on the heterogeneous disk environment. But it is difficult to do so because the performance and capacity of each disk are different. We think to extend AOD method for this aim. This method uses the difference of access load between primary data and backup data and balance access load and data amount simultaneously on the homogeneous environment. In this paper, we consider to maximize usable capacity by the position of backup and the disk sequencing. Furthermore, we arrange the primary data in order to balance the performance based on its Bandwidth to Space Ratio.

In recent years, the method of assigning labels to the nodes of an XML tree is getting more attraction. Various functions in an RDBMS can be easily utilized by storing the labeled XML documents into the RDB. However, in traditional labeling methods, a number of nodes need to be relabeled, when the XML documents are updated. To address this problem, we proposed DO-VLEI code combining VLEI code with the Dewey Order method. DO-VLEI code is effective to reduce the update cost, but the label size increases rapidly when handling large XML documents. To reduce the label size, we presented Compressed-bit-string DO-VLEI (C-DO-VLEI) code. However, it is difficult to handle the length of C-DO-VLEI because it is a variable-length code. In this paper, we propose two effective methods, VLEI-ABL and VLEI-EOL for handling the code length of C-DO-VLEI. We perform experiments to compare the storage consumption of the proposed methods with the previously known ORDPATH. The experimental results show that our methods considerably outperform the ORDPATH.

The objective of our research has been to achieve energy-saving key-value store system. We introduce architecture of our distributed key value store and control framework for energy saving. To reduce power consumption, the control framework should reduce the time required before node power-down. Since this time depends heavily on the amount of data migration, we employ an approach which reduce that amount. We describe and evaluate three control strategies that can be executed on this control framework, each consisting of a data allocation algorithm and data control layout algorithm for use when nodes are to be powered down. With a prototype system developed for evaluation purposes, the time required before node power down is reduced by use of our proposed data allocation algorithm, by means of its taking into account the possible presence of Write off-loading nodes.

Many recent advanced applications of information systems require large amounts of data to be stored in multiple hard disk drives (HDDs). The placement of data in these drives is quite influential in reducing energy consumption while retaining the performance of the system. Several approaches have been proposed in which frequently accessed files are concentrated into a limited number of drives to enable spin-down of the other drives. However, the placement of infrequently accessed data in these drives is also significant because energy consumption during spin-up to access these data cannot be ignored, particularly when files tending to be used together are placed on many different drives. In this paper, we propose a novel method named PLECO (Placement of files for Latency and Energy Consumption Optimization), which aims to place correlated files into the same drive to reduce power consumption and improve the performance of the system. We simulated and evaluated the proposed method, and these results indicate that our proposal can reduce both the energy consumption and the access latency by up to 32% and 92%, respectively, compared with a baseline system.

Every day, information on the Web becomes increasingly enriched. Web access is now very useful in many aspects of daily life, particularly for writing documents and programs. In fact, it has become quite usual to edit files while referring to information on the Web. During the file-editing process, we usually visit so many Web pages that we cannot remember all of the relevant ones. Later, if we want to revisit the same Web pages to modify some part of a file, it can be very hard to track down the Web pages originally referred to. In this paper, we propose methods for finding relationships between files and Web pages based on the co-occurrence of data in Web-access logs and file-access logs. These relationships are very useful for revisiting Web pages related to target files. To analyze co-occurrence in these two types of access logs, there are two approaches for merging the logs, involving a trade-off between accuracy and execution time. We call them the Pre-Merge and Post-Merge methods, and we have evaluated these two methods using actual access logs.

Recently, the researchers have proposed a number of labeling schemes. In these labeling schemes, the approach which can extract structural information between nodes and process query efficiently is more outstanding. However, most of these labeling schemes do not well support update operations. To achieve update-friendly operations, some of the methods keep intervals between labeling numbers, but it requires whole relabeling when the intervals are used up. Several labeling schemes support dynamic XML documents, but most of these labeling schemes allow only leaf node insertions. OrdPathX supports both leaf node insertions and internal node insertions. Inspired by the method of inserting internal nodes of OrdPathX and extending the C-DO-VLEI code, in this paper we propose two dimensions VLEI code. We discuss how this labeling scheme labels nodes and how we can get the structural information of nodes from their labels. We design experiments to evaluate the efficiency of producing labels, the storage consumption and the querying performance of two dimensions VLEI code we proposed, and compare those with the OrdPathX.

In this paper, we we address the data access efficiency and scalability by introducing a first trial of combining parallel B-tree structures with open-source database management systems PostgreSQL. Because the high cash hit rates and low update synchronization cost in the Fat-Btree index, our DBMS has a good scalability and data accessing efficiency. In addition, the center node is no more required for the metadata management in the system, thus our DBMS achieves high reliability as well. To evaluate our system, we compare it with HBase, a distribute DBMS built on top of Hadoop, which is known for the efficient and scalable data accessing performance. Experimental results on a 100-nodes cluster system verify that our DBMS greatly outperform HBase in both system scalability and throughput.

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.

When multiple users access a network storage system for cloud computing, security becomes a key factor in the service, as well as performance and reliability. The "encrypt-on-disk'' scheme effectively protects transmitted and stored data in network storage. However, this scheme has the problem of revocation for shared files. Active revocation is safe but has denial periods to allow immediate reencryption, while lazy revocation has no denial period but is unsafe during the delay. We propose intelligent storage nodes capable of handling active revocation in storage without the denial period by adopting a primary--backup configuration. This approach provides a good combination of security and availability by replication. However, the reencryption process negatively affects the update performance. Delaying the reencryption process and disk write on the backup node improves performance with no ill effect on security and a small decrease of MTTDL for the simple primary-backup configuration. We evaluate the performance of the proposed approaches by experiments, and the reliability by estimation.

Recently, labeling methods to extract and reconstruct the structural information of XML data, which are important for many applications such as XPath query and keyword search, are becoming more attractive. To achieve efficient structural information extraction, in this paper we propose C-DO-VLEI code, a novel update-friendly bit-vector encoding scheme, based on register-length bit operations combining with the properties of Dewey Order numbers, which cannot be implemented in other relevant existing schemes such as ORDPATH. Meanwhile, the proposed method also achieves lower storage consumption because it does not require either prefix schema or any reserved codes for node insertion. We performed experiments to evaluate and compare the performance and storage consumption of the proposed method with those of the ORDPATH method. Experimental results show that the execution times for extracting depth information and parent node labels using the C-DO-VLEI code are about 25% and 15% less, respectively, and the average label size using the C-DO-VLEI code is about 24% smaller, comparing with ORDPATH.

Traditional full-text searches cannot search keyword-lacking files, even if the files are related to the keywords. In this paper, we propose a method for searching keyword-lacking files named FRIDAL (File Retrieval by Interfile relationships Derived from Access Logs). The proposed method derives interfile relationship information from file access logs in the file server, based on the concept that those files opened by a user in a particular time period are related.

We have proposed a method replica-assisted migration that use temporally replica data in other storage nodes during executing data migration to keep both qualities of service accesses and data migration duration. However, we have not evaluated the method with real workload including update operations. In this paper, we first improve the method and evaluate an efficiency of the method under file server workloads. Base on the result of it, we also propose adaptive replica-assisted migration for reduce migration duration.

The secure storage systems adopting the encrypt-on-disk scheme, in which files are stored in cipher for efficient data transmission, need to re-encrypt files with new cryptographic keys when a revocation occurs. There are two re-encryption methods, namely active revocation in which the re-encryption is immediately performed and lazy revocation in which the re-encryption is delayed until the file is updated. There is the trade-off between performance and security because active revocation has the expense of immediate re-encryption, while lazy revocation is vulnerable during its re-encryption delay. We consider the environment in which re-encrypted file is pre-computed by using backup data in a parallel storage system effective for this issue. However, the performance of update is decreased on account of the difference of keys used in primary and backup. In this paper, we evaluate a method in which the differential data re-encrypted for backup are not written to the file but be kept on the memory in different key environment, and evaluate the different key environment in parallel storage by experiment.

"null", More and more within the last couple of years, there is a recognition that the Web services composition concept will constitute a major breakthrough and will revolutionize the way we deal with integrating disparate and distributed computing environments. Yet, to rise to such a position, a chief concern is to guarantee a high-dependability level of the Web services compositions, which is significantly critical, specially in view of the particularities of the Web services environment( unpredictability, heterogeneity, autonomy), if confronted with other computing environments. Our present work falls within this context since we tackle the problem of QoS (Quality of Service) in the Web services context by verifying to what extent fault-tolerant and dynamically-executed Web services compositions are effi- ciently serving their purposes. And in pursuing this goal, we introduce a novel model that characterizes, estimates and analyzes several QoS properties of dynamically-executed fault-tolerant Web services compositionsnamely the reliability and the execution time. Our model allows acquiring more accurate estimations since it confers a paramount importance to the repercussions of failures. In addition, contrary to other QoS estimations models in the Web services context, which use QoS estimations published in UDDIs by the Web Services owners/ providers, our model computes QoS estimations on the base of the compositions execution observations, where the observation results are collected in a history. Finally, since Web services are stateless, tracking the failures and determining their locations is almost impossible. To overcome this limitation, we propose to attach to each of the composition’s component a state. In doing so, obtained estimations can contribute in acquiring more accurate information about the failures locations and can be used later to improve the composition QoS in the future.

A new concurrency control protocol for parallel Btree structures, MARK-OPT, is proposed. MARK-OPT marks the lowest structure-modification-operation (SMO) occurrence point during optimistic latch coupling operations, to reduce the cost of SMO compared to the conventional protocols such as ARIES/IM and INC-OPT. The marking reduces the frequency of restarts for spreading the range of X latches, which will clearly improves the system throughput. Moreover, the MARK-OPT is deadlock free and satis- fies the physical consistency requirement for Btrees. These indicate that the MARK-OPT is right and suitable as a concurrency control protocol for Btree structures. This paper also proposes three variations of the protocol, INC-MARKOPT, 2P-INT-MARK-OPT and 2P-REP-MARK-OPT, by focusing on tree structure changes from other transactions. We implement the proposed protocols, the INC-OPT, and the ARIES/IM for the Fat-Btree, a form of parallel Btree, and compare the performance of these protocols using a large-scale blade system. The experimental results indicate that the proposed protocols always improve the system throughput, and the 2P-REP-MARK-OPT is the most useful protocol in a high update environment. Moreover, in the experiment, the low frequency of restarts in the proposed protocols indicates that the marking in the proposed protocols is effective.

We have previously proposed a unified presentation contents search mechanism named UPRISE (Unified Presentation Slide Retrieval by Impression Search Engine), and have also proposed a method to use laser pointer information in lecture scene retrieval. In this paper, we discuss the treatment of the laser pointer and speech information, and propose two methods to filter the laser pointer information using keyword occurrence in slides and speech. We also propose weighting schemata with filtered laser pointer information using slide text and speech information. We evaluate our approach by using actual lecture videos and presentation slides.