Tomohiro Yoshihara

Synthesis of Cu2ZnSn(S,Se)4 Nanoparticles for Application in Low-Cost Solar Cells

A low-temperature colloid synthesis approach was applied to synthesize Cu2ZnSn(S,Se)4 (CZTSSe) nanoparticles for the first time. This method produced gram quantities of material with a chemical yield in excess of 90% within a short synthesis time. Average sizes of 80 nm nanoparticles with narrow diameter distribution were achieved using Na2S with Na2Se as a chalcogenide source. The formation reaction of kesterite CZTSSe fabricated from synthesized nanoparticles was investigated, and the results showed that the sufficiently high partial pressures of Sn with Se during the annealing process were important for the fabrication of high-quality CZTSSe films.

A concurrency control protocol for parallel B-tree structures without latch-coupling for explosively growing digital content

While shared-nothing parallel infrastructures provide fast processing of explosively growing digital content, managing data efficiently across multiple nodes is important. The value-range partitioning method with parallel B-tree structures in a shared-nothing environment is an efficient approach for handling large amounts of data. To handle large amounts of data, it is also important to provide an efficient concurrency control protocol for the parallel B-tree. Many studies have proposed concurrency control protocols for B-trees, which use latch-coupling. None of these studies has considered that latch-coupling contains a performance bottleneck of sending of messages between processing elements (PEs) in distributed environments because latch-coupling is efficient for a B-tree on a single machine. The only protocol without latch-coupling is the B-link algorithm, but it is difficult to use the B-link algorithm directly on an entire parallel B-tree structure because it is necessary to guarantee the consistency of the side pointers. We propose a new concurrency control protocol named LCFB that requires no latch-coupling in optimistic processes. LCFB reduces the amount of communication between PEs during a B-tree traversal. To detect access path errors in the LCFB protocol caused by removal of latch-coupling, we assign boundary values to each index page. Because a page split may cause page deletion in a Fat-Btree, we also propose an effective method for handling page deletions without latch-coupling. We then combine LCFB with the B-link algorithm within each PE to reduce the cost of Structure Modification Operations (SMOs) in a PE, as a solution to the difficulty of consistency management for the side pointers in a parallel B-tree structure. To compare the performance of the proposed protocol with conventional protocols MARK-OPT, INC-OPT, and ARIES/IM, we implemented them on an autonomous disk system with a Fat-Btree structure. Experimental results in various environments indicate that the system throughput of the proposed protocols is always superior to those of the other protocols, especially in large-scale configurations, and LCFB with the B-link algorithm is effective at higher update ratios.

MARK-OPT: A Concurrency Control Protocol for Parallel B-Tree Structures to Reduce the Cost of SMOs

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.

An Efficient Commit Protocol Exploiting Primary-Backup Placement in a Distributed Storage System

Advanced data engineering applications require a large-scale 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 that use a primary-backup 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 are 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

Evaluation of Placement and Access Asignment for Replicated Object Striping

The number of stored objects that should be targets of high throughput retrieval, such as multimedia stream objects, is increasing recently. To implement a high throughput storage system, striping technique using multiple disk drives are commonly used. However, the ordinary disk striping methods implemented in RAID 0, 3-5 have problems of the flexibility, extensibility, and quality of services (QoS). We have proposed the autonomous disk cluster to realize the flexible and extensible storage system by treating each target object as a unit for management. It also adopts the primary-backup technique for the object to satisfy required QoS. We have shown that the autonomous management for the object unit is effective. However, there is room for improving its throughput. In this paper, we consider an approach for importing a flexible striping technique into the autonomous disk system to provide high throughput. It is important to locate fragments of an object into a part of disk cluster appropriately with considering workload skews to derive the required throughput. We use access history to decide their location. We also propose a method to divide access load into fragmented primary and backup adaptively. The experimental results using PC cluster indicate that the proposed methods are effective.

A Concurrency Control Method for Parallel Btree Structures

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.

A Concurrency Control Protocol that Selects Accessible Replicated Pages to Avoid Latch Collisions for B-Trees in Manycore Environments

In recent years, microprocessor vendors aiming for dramatic performance improvement have introduced manycore processors with over 100 cores on a single chip. To take advantage of this in database and storage systems, it is necessary for B-trees and their concurrency control to reduce the number of latch collisions and interactions among the cores. Concurrency control methods such as physiological partitioning (PLP), which assigns cores to partitions in a value–range partition, have been studied. These methods perform effectively for nearly static and uniform workloads using multicore processors. However, their performance deteriorates significantly if there is major restructuring of B-trees against skew and for changing workloads. The manycore approach has a high likelihood of causing workload skew, given the lower power of each core, with an accompanying severe degradation in performance. This issue is critical for database and storage systems, which demand consistent high performance even against dynamic workloads. To address this problem, we propose an efficient new concurrency control method suitable for manycore processor platforms, called the selecting accessible replicated pages (SARP) B-tree concurrency control method. SARP achieves a consistent high performance with robustness against workload skew by distributing the workload to many cores on manycore processors, while reducing latch collisions and interactions among the cores. By applying parallel B-trees to shared-everything environments, SARP selects execution cores and access paths that distribute workloads widely to cores with appropriate processor characteristics. Experimental results using a Linux server with an Intel Xeon Phi manycore processor demonstrated that the proposed system could achieve a throughput of 44 times that for PLP in the maximum-skew case and could maintain the throughput at 66% of a throughput for uniform workloads.

Restorable Update Write by SSD Controller to Improve Destage Throughput in Enterprise Storage System

RAID5/6 is widely adopted in enterprise storage systems, despite having a write-penalty problem and consuming the cache memory bandwidth. One of the existing ways for reducing data transfer overhead using XDWRITE/XPWRITE determined in SCSI specification. However, this method still consumes bandwidth because of duplicating the XOR data into cache memories for failure recovery. In this paper, we propose a new destaging method that transfers the XOR data through the buffer memory. In addition, we propose new commands that provide a roll back function exploiting the FTL in SSDs. It simplifies the failure recovery in the case of a buffer memory failure while destasing. We evaluated the throughput improvement using an analytical model. Our proposal improved the Random Write throughput by 33% and SPC Benchmark-1™ throughput by 29%, from the conventional method.

a-SiGeC:H solar cells fabricated near the threshold of the amorphous-to-crystalline transition for narrow gap solar cells and its improvement by modifying interfaces

a-SiGeC:H thin films have been deposited over a broad range of MMG/SiH<inf>4</inf> and H<inf>2</inf>/SiH<inf>4</inf> gas flow ratios. It was observed that the optical band-gap (E<inf>opt</inf>) of the films was gradually reduced as either MMG/SiH<inf>4</inf> or H<inf>2</inf>/SiH<inf>4</inf> increases. Then, a-SiGeC:H solar cells containing less carbon were fabricated near the threshold of amorphous-to-crystalline transition for the fabrication of narrow-gap solar cells, because high H<inf>2</inf> dilution is beneficial for suppressing C incorporation. It was found that, at the fixed MMG/SiH<inf>4</inf>, the solar cells showed the improving trends in performance under higher H<inf>2</inf> dilution, if the a-SiGeC:H thin films did not become crystalline. The optimized conditions were obtained under much higher H<inf>2</inf> dilution when MMG flow rates were higher. From the QE measurement under negative bias voltage, it was assumed that the origin of the deterioration of the cell performance was stemmed from the reduced mobility-lifetime (μτ) product due to the presence of C and Ge. The various techniques were employed in order to improve the performance mainly by modifying interface property rather than i-layer itself. As a result, the solar cells were notably improved.