Abstract: A modularized computing system includes a plurality of modular components that are coupled together forming a network. Each modular component includes a standard network interface. The system further includes an initialization module, a monitor module, a storage medium, and a management module. As a module unit is coupled to the network, the initialization module automatically configures the component to an operable state. The monitor module monitors network operations including performance parameters of each modular component based on a plurality of system policies. Based on information gathered by the monitor module, the management module actively modifies network structure and resource allocation to optimize network performance.

Abstract: A modularized computing system includes a plurality of modular components that are coupled together forming a network. Each modular component includes a standard network interface. The system further includes an initialization module, a monitor module, a storage medium, and a management module. As a module unit is coupled to the network, the initialization module automatically configures the component to an operable state. The monitor module monitors network operations including performance parameters of each modular component based on a plurality of system policies. Based on information gathered by the monitor module, the management module actively modifies network structure and resource allocation to optimize network performance.

Abstract: The present invention is directed toward a data management appliance and ancillary technologies for replicating data written to a primary storage system. The data management appliance is a random-access storage system that at the logical block level replicates the contents of a primary storage system over time.

Abstract: The present invention is an interprocess communications (IPC) system and method which utilizes shared memory to transfer message contents and a separate, efficient notification mechanism to perform message control functions. The present invention supports robust, asynchronous communications between potentially large number of software application processes residing on any single or tightly-coupled multiple processor arrangement supporting shared memory between the processes. The present invention is a combination of existing IPC mechanisms available in most commercially available UNIX operating systems. The IPC mechanism utilizes the presently available UNIX shared memory IPC mechanism for transferring message contents and the presently available UNIX datagram socket IPC mechanism for message control and notification. The IPC mechanism may be adapted to other operating system environments.

Abstract: A bar code conversion system and method identifies good data in the scan data array and analyzes that data to generate machine-interpretable code. A vision system optically scans bar code labels and places the resulting digital camera data into a two dimensional line scan data array of pixels. A process.sub.-- line function then searches the scan data array for three lines that intersect the initial bars and spaces of the bar code (called starting lines) and 3 lines that intersect the last bars and spaces (called ending lines). For each pair of starting and ending lines, a build.sub.-- composite generates a composite line using pixels in positions between the first pixel of a starting line and the last pixel of a corresponding ending line. A run.sub.-- length decode function is then invoked on each composite line to create a run length array of width values, each of which indicates the width (in pixels) of a bar or space of the bar code. Using statistical analysis, a convert.sub.-- bars.sub.-- and.sub.

Abstract: A method and apparatus are described for on-the-fly cataloging of library cell contents in an automated robotic tape library. This method decouples the robotic arm motion/tape cartridge label scanning process from the image processing software/hardware. In accordance with the present invention, tape cartridge label images are captured at a library system processor priority which is higher than the priority of the image processing task. These captured images are stored in a circular buffer while image processing continues to take place in a lower priority task. Both the image processing system hardware and software are asynchronous with respect to the velocity of the robotic arm used to scan the tape cartridge cells. The scanning velocity of the robotic arm can be maximized with respect to purely mechanical, electrical, or optical camera considerations and is independent of the library system processor image processing speed.