International Telemetering Conference Proceedings / October 21, 2002 / Town & Country Hotel and Conference Center, San Diego, California / This paper presents efforts in developing a data management system and storage infrastructure for assisting test engineers in achieving information superiority and maintaining vital up-to-date information. The focus of this Paper is to generate support for a technology refresh, upgrading the major data centers that share in the responsibility of processing telemetry information. We illustrate how our efforts fit into this goal and provide an overview of our concept for a revolutionary transformation in data management systems. We present the significance of this new technology and suggest a path to implementing the solution.

No description available.

No description available.

International Telemetering Conference Proceedings / October 25-28, 1993 / Riviera Hotel and Convention Center, Las Vegas, Nevada / With the development of microelectronics and computer technology, telemetry computer systems are demanded to provide larger storage capacity and higher storage data rate than ever before. This paper fully considers various factors of a high-speed PCM fiber-optic telemetry system such as data format, data rate, data storage, the width of data storage, storage data rate. All these considerations lead to a new scheme with a semiconductor RAM and a dedicated program as its basic idea. This scheme chooses 1Mbits or 4Mbits static-RAM chips to implement the telemetry data storage device with a total capacity of 4Mbytes, 16Mbytes, or 64Mbytes. The software running on COMPAQ 386/25M or its compatibles is written in Turbo C 2. 0 to fetch, decompose, display and process data stored in the large-capacity RAM. The main task of the system processing software is to identify the flag words of frame sync-code -pattern and then demultiplex the data into separate channel data to be stored in the disk. Besides the ability to recognize specific data format, the software can also rectify data confusion to some extent. The scheme has already been proved to be efficient to receive large capacity of data with features of high data rate, high data storage in a short time.

Current hard drive technology shows a widening gap between the ability to store vast amounts of data and the ability to process. To overcome the problems of this secular trend, we explore the use of available distributed RAM resources to effectively replace a mechanical hard drive. The essential approach is a distributed Linux block device that spreads its blocks throughout spare RAM on a cluster and transfers blocks using network capacity. The presented solution is LAN-scalable, easy to deploy, and faster than a commodity hard drive. The specific driving problem is I/O intensive applications, particularly digital forensics. The prototype implementation is a Linux 2.4 kernel module, and connects to Unix based clients. It features an adaptive prefetching scheme that seizes future data blocks for each read request. We present experimental results based on generic benchmarks as well as digital forensic applications that demonstrate significant performance gains over commodity hard drives.

Photoactivated fluorescence from individual silver nanoclusters ranging in size from 2 8 atoms has been demonstrated at room temperature. The optical properties of such clusters are far superior to those of fluorescence dyes with absorption cross sections ~50 times stronger than those of even the best organic dyes. The strong oscillator strengths produced from such nanoclusters has been shown to yield comparable enhancement factors in the surface-enhanced Raman spectroscopy (SERS) process to those observed in the presence of a plasmon- supporting nanoparticle. Raman transitions are in fact so strong that antistokes scattering is also observable on a single molecule (SM) level marking the first true demonstration of SM-SERS to date. Capable of generating true scaffold specific Raman scattering on the single molecule level, the combination of fluorescence from the small nanoclusters and strong observed Raman signals in the absence of a nanoparticle strongly indicate a chemical or charge transfer SERS enhancement mechanism.

No description available.

No description available.

This thesis documents a research project that has led to the design and prototype implementation of several data storage schemes suited to the efficient multi-scale representation of integrated spatial data. Spatial information systems will benefit from having data models which allow for data to be viewed and analysed at various levels of detail, while the integration of data from different sources will lead to a more accurate representation of reality. The work has addressed two specific problems. The first concerns the design of an integrated multi-scale data model suited for use within Geographical Information Systems. This has led to the development of two data models, each of which allow for the integration of terrain data and topographic data at multiple levels of detail. The models are based on a combination of adapted versions of three previous data structures, namely, the constrained Delaunay pyramid, the line generalisation tree and the fixed grid. The second specific problem addressed in this thesis has been the development of an integrated multi-scale 3-D geological data model, for use within a Geoscientific Information System. This has resulted in a data storage scheme which enables the integration of terrain data, geological outcrop data and borehole data at various levels of detail. The thesis also presents details of prototype database implementations of each of the new data storage schemes. These implementations have served to demonstrate the feasibility and benefits of an integrated multi-scale approach. The research has also brought to light some areas that will need further research before fully functional systems are produced. The final chapter contains, in addition to conclusions made as a result of the research to date, a summary of some of these areas that require future work.

As the size of cloud systems and the number of hosted virtual machines (VMs) rapidly grow, the scalability of shared VM storage systems becomes a serious issue. Client-side flash-based caching has the potential to improve the performance of cloud VM storage by employing flash storage available on the VM hosts to exploit the locality inherent in VM IOs. However, there are several challenges to the effective use of flash caching in cloud systems. First, cache configurations such as size, write policy, metadata persistency and RAID level have significant impacts on flash caching. Second, the typical capacity of flash devices is limited compared to the dataset size of consolidated VMs. Finally, flash devices wear out and face serious endurance issues which are aggravated by the use for caching. This dissertation presents the research for addressing these problems of cloud flash caching in the following three aspects. First, it presents a thorough study of different cache configurations including a new cache-optimized RAID configuration using a large amount of long-term traces collected from real-world public and private clouds. Second, it studies an on-demand flash cache management solution for meeting VM cache demands and minimizing device wear-out. It uses a new cache demand model Reuse Working Set (RWS) to capture the data with good temporal locality, and uses the RWS size (RWSS) to model a workload?s cache demand. Finally, to handle situations where a cache is insufficient for VMs? demands, it employs dynamic cache migration to balance cache load across hosts by live migrating cached data along with the VMs. The results show that the cache-optimized RAID improves performance by 137% without sacrificing reliability, compared to traditional RAID. The RWSS-based on-demand cache allocation reduces workload?s cache usage by 78% and lowers the amount of writes sent to cache device by 40%, compared to traditional working set based cache allocation. Combining on-demand cache allocation with dynamic cache migration for 12 concurrent VMs, results show 28% higher hit ratio and 28% lower 90th percentile IO latency, compared to the case without cache allocation.