Hewlett-packard Development Company, L.p.

Klaus Wurster

Pankaj K Garg

Umeshwar Dayal

Ming C Hao

Vijay Machiraju

Peter J Wright

Kave Eshghi

Ming-Chien Shan

Fabio Casati

Jerremy Holland

Mohamed E Dekhil

Martin L Griss

Thomas Gschwind

Stephany Kalil

Betty Cant

A graphical production system produces pixel bar charts, in various embodiments. In an embodiment, a pixel bar chart is constructed that may include multiple bars with substantially equal widths and potentially variable heights to enable visual data analysis. A bar may include a number of pixel units corresponding to a volume of data records represented within the bar. Selected visual indicators are displayed within the pixel units of the bar, where a selected visual indicator represents a value of an attribute of a data record represented by the pixel unit. In an embodiment, the pixel units may be ordered within a bar in a particular sequence (e.g., ascending, descending or some other order). Further, in an embodiment, the bar width is determined to minimize the amount of padding. In an embodiment, a visual indicator scale is displayed to indicate values associated with each visual indicator.

Constructing substantially equal-width pixel bar charts to enable visual data analysis

A method for generating a pixel-oriented graph provides visual boundaries for representing aggregate information. The present method for generating a pixel-oriented graph includes determining a visual boundary for representing an aggregate of a set of information depicted in the pixel-oriented graph and constructing a set of pixel blocks that represent the information such that the pixel blocks are visually distinguished by the visual boundary. A method according to the present techniques employs a visual weight to classify labels and high weight information is shown and other information is made invisible and hidden in the property of a graph.

Visual boundaries for aggregate information in pixel-oriented graphs

A method and system for tracking hits of a requested network file includes embedding an instruction within the file that directs the requesting device to transmit an indicator to a processor for counting subsequent to the requesting device receiving the requested file. The request for the network file is made by a requesting device (e.g., client of an end-user) and may be made over the Internet. The instruction is executed by a Web browser at the requesting device. The instruction is embedded within the Internet file, so that if the file is cached by a proxy server, the instruction is also cached. Thus, the instruction will reach the end-user regardless of whether the file is downloaded from the proxy server or from an originating server. When the file reaches the end-user, the instruction triggers the transmission of the indicator that is send to the processor for counting.

Tracking hits for network files using transmitted counter instructions

An auto-discovery system discovers related components of a service operating on a distributed network. This auto-discovery system involves (1) selecting a model of the configuration of the service; (2) determining, from the model, &ldquo;scope hints&rdquo; that limit the possible locations of related components; (3) determining a root component of the service; (4) using scope hints to limit the search, to rapidly find related instances of components contained in the root service component; (5) determining whether a quorum of components comprising a distributed service has been discovered; (6) determining the parent service component of each discovered service component by instance matching; and (7) using the instance match information to determine the discovery tree of components that have the structure of the service map, share a common root component, and define a service group. This auto-discovery system efficiently discovers the components of a wide range of services within a distributed environment. The system limits the scope of searches for components related to previously discovered components. The system determines the relationships between the discovered service components, and matches them into groups of related services. This enables an operational manager of a distributed service to better determine service configurations, expand resource bottlenecks, and improve an end-users' experience.

Method and apparatus auto-discovering components of distributed services

Growing community of inventors

Austin, TX, United States of America

Klaus Wurster