Jingli Zhou

Design and implementation of a data disaster recovery system based on storage-virtualization

In this paper, we design and implement a data disaster recovery system based on storage-virtualization. The data disaster recovery system implements its function through the collaboration of multiple sites in wide-area scope. Each site is an independent data center and data disaster recovery unit. The construction of each site is based on storage-virtualization gateway which is a storage-virtualization platform supporting storage multi-protocol. Therefore, through storagevirtualization gateways implementation of data replication, failure detection, service switch/takeover among the multiple sites, data disaster recovery could be implemented.

The building strategy of iSCSI appliance

Even though the foundation of iSCSI (Internet SCSI) protocol is quite different from that of Fibre Channel, and maybe the original idea of replacing FC-SAN with truly IP-based iSCSI-SAN is hard to be fulfilled in short term, nobody can ignore the broadly existence and the bright future of the iSCSI technology. We present an introduction to build several iSCSI appliances in this paper from a single stand-alone iSCSI Target, an iManager-based iSCSI appliance, and our embedded iSCSI Target implementation to a RAIN (Redundant Array of Independent Nodes) -based iSCSI appliance. Not only do we offer some experiment results of our standalone iSCSI Target and our embedded iSCSI Target, our in-band and out-of-band iManager-based iSCSI Target designing strategy and performance comparison, but also give an explanation of our RAIN-based iSCSI Target implementation concerning to the topology of our RAIN Target, the node data distribution method and our ways to deal with the capacity-expansion-on-demand problem.

Modeling general AIMD congestion control throughput with slow start

General Additive Increase and Multiplicative Decrease (General AIMD or GAIMD) congestion control generalizes the standard TCP congestion control mechanism. In this paper, we present GAIMD-SS, an enhanced model for predicting the long-term steady-state mean throughput of the GAIMD congestion control, towards more accurate results than previous GAIMD model. We develop a three-state Markov chain for analyzing the behaviors of the GAIMD, and enhance previous works by taking into account the slow start phase and receivers maximum window limitation. Our experiment and simulation results show that the GAIMD-SS model predicts the sending rate of GAIMD congestion control more accurately than the previous works in a wider range of packet loss rate.

An implementation of iSCSI HBA based on Intel IOP80321

Generally speaking, the iSCSI initiator function can be fulfilled in about 3 different ways, they are software iSCSI for various OSs (operating systems), software iSCSI with TCP/IP offload engine (TOE) and hardware iSCSI with TCP/IP offloading, and the third one can be regarded as an iSCSI HBA (Host Bus Adapter). Based upon Intelligent I/O (I2O) Architecture Specification and Intel(R) IQ80321 I/O processor evaluation board (PEB), a prototype of an iSCSI HBA implementation is presented in this paper. An OSM (OS Service Module) design introduction under Linux OS is put forward, and an embedded Linux DDM (Device Driver Module) development is also explained. By comparing with both the Intel(R) PRO 1000T IP storage NIC and Microsoft iSCSI Initiator version 1.06, some testing analysis of this iSCSI HBA in the 100/1000Gbps Ethernet network circumstance are given in the end.

Analysis and amendment of flow control credit-based in SAN extension

As organizations increasingly face an enormous influx of data that must be stored, protected, backed up and replicated. One of the best ways to achieve the goal is to interconnect geographically dispersed SANs through reliable and high-speed links. In this storage extension application, flow control deals with the problem where a device receives the frames faster than it can process them, when this happens, the result is that the device is forced to drop some of the frames. The FC flow control protocol is a credit-based mechanism and usually used for SAN extension over WDM and over SONET/SDH. With FC flow control, when a source storage device intends to send data to a target storage device, the initiating storage device must receive credits from target device. For every credit the initiating device obtains, it is permitted to transmit a FC frame, so congestion is always avoided in the network. This paper analysis the mechanisms of FC flow control and its limitation in SAN extension when the extension distance increases. Computing result indicates that the maximum link efficiency and throughput in SAN extension have relation to credits, frame size and extension distance. In order to achieve the maximum link efficiency and throughput, an extended FC flow control mechanisms are proposed.

Designing and application of SAN extension interface based on CWDM

As Fibre Channel (FC) becomes the protocol of choice within corporate data centers, enterprises are increasingly deploying SANs in their data central. In order to mitigate the risk of losing data and improve the availability of data, more and more enterprises are increasingly adopting storage extension technologies to replicate their business critical data to a secondary site. Transmitting this information over distance requires a carrier grade environment with zero data loss, scalable throughput, low jitter, high security and ability to travel long distance. To address this business requirements, there are three basic architectures for storage extension, they are Storage over Internet Protocol, Storage over Synchronous Optical Network/Synchronous Digital Hierarchy (SONET/SDH) and Storage over Dense Wavelength Division Multiplexing (DWDM). Each approach varies in functionality, complexity, cost, scalability, security, availability , predictable behavior (bandwidth, jitter, latency) and multiple carrier limitations. Compared with these connectiviy technologies,Coarse Wavelength Division Multiplexing (CWDM) is a Simplified, Low Cost and High Performance connectivity solutions for enterprises to deploy their storage extension. In this paper, we design a storage extension connectivity over CWDM and test its electrical characteristic and random read and write performance of disk array through the CWDM connectivity, testing result show us that the performance of the connectivity over CWDM is acceptable. Furthermore, we propose three kinds of network architecture of SAN extension based on CWDM interface. Finally the credit-Based flow control mechanism of FC, and the relationship between credits and extension distance is analyzed.

Real-time multimedia transmission over Bluetooth wireless Internet based on differentiated services

Supporting end-to-end QoS of media streaming applications over wireless Internet has received tremendous interest recently. In this paper we propose a new architecture for transmitting real-time multimedia streams over Bluetooth wireless Internet integrated with Differentiated Services (DiffServ) mechanism. IP support over Bluetooth is discussed in details to take advantage of the DiffServ framework. In our proposed architecture, the Bluetooth wireless users are connected to the Bluetooth LAN Access Point to acquire the capability of Internet access. Different media streams are classified at edge routers of the DiffServ core network, mapped to different priorities and delivered according to the dynamically negotiated Service Level Agreement (SLA). QoS monitoring agents deployed in network nodes and endsystems can monitor network QoS conditions and provide reports to the concerned nodes. Implementation details of the prototype of the proposed architecture are given in this paper. Specially, the Bluetooth communication is implemented as a virtual network card that supports IP protocol in order to provide consistent access to the DiffServ Internet. Experimental results show that enhanced end-to-end QoS can be achieved based on our architecture.

New lossless precompression algorithm suitable for CMOS digital camera

CMOS image sensor is the enable technology for videoconferencing over PSTN going to public. The major advantage of CMOS image sensor over traditional CCD is that CMOS sensor can integrate more circuits on the same sensor chip. In this paper, we proposed a new lossless pre- compression algorithm targeted for CMOS integration. Our pre-compression algorithm gains lossless compression ratio about 4 and there is no multiply and division operation required. We evaluate the performance of this algorithm in this paper and introduce a hardware architecture that can be easily integrated into CMOS image sensor chip.

Fractal-based wavelet transform coding for low-bit-rate video

Fractal image compression allows fast decoding but suffers from long encoding time. By mapping similar regions within an image in the wavelet domain, we can use biorthogonal wavelet transform coefficients to find mappings which are identical to conventional blockwise mappings,thus the time consuming encoding process will improved by wavelet based hierarchical interpretation. We apply our methods to encode video sequence of digital image, in our scheme, the encoding process ID done for image pieces or motion vectors from the predecessor. The result shows a relatively low frame rate, since our target is transmitting video sequence in the level of requirements in ITU-T H.324 or the V.34 modem physical layer.

Testing of a magnetic strip with grey predicting control

Along with the spreading of the magnetic cards in our country, it is necessary to develop new equipments and methods for manufacture and test of the magnetic card to ensure that the magnetic card meet the requirements of the ISO standards. The testing system will worked out do all the test items to the magnetic cards according to the related test standards made by the ISO TC/97 group. Considering the actuall testing conditionin factory , we devided the magnetic characteristics testing into special test and general test In this paper, the hardwire design and configuation. software functions . working mode and characteristic of the system are discussed in detail.