IBM Autonomic Task Manager for Administrators (ATMA) is a spreadsheet-based scripting environment for quickly composing and automating system management tasks. With this environment, administrators can execute management commands and combine these commands to create ad hoc scripts and visualizations of system management information. The basic building blocks for these tasks are spreadsheet templates that are customized with a simple drag-and-drop interface. Autonomic Task Manager for Administrators enables the insertion of GUI, visualization, or system management components into cells of a spreadsheet and customization of the cells in order to insert control logic for a system management solution. Just as in spreadsheets, data in the cells are automatically processed and updated; this feature allows real-time system data feeds.

Currently, Autonomic Task Manager for Administrators supports a variety of system management plug-ins, including Simple Network Management Protocol (SNMP), Secure Shell (SSH), and Java™ Management Extension (JMX). Using ATMA's component plug-in API, developers can build custom components that can be used to develop tools using different management APIs; one such API interfaces to IBM Autonomic Integrated Runtime Environment, which allows communication with resources based on Web Services Distributed Management (WSDM). ATMA can also interact with any Java object.

Autonomic Task Manager for Administrators significantly reduces script creation time with its familiar spreadsheet interface and building blocks made from templates. The package includes the executable, relevant plug-ins, installation instructions, and user documentation.

How does it work?

System administrators or value-added re-sellers (VARs) can use the spreadsheet-based scripting environment to build scripts incrementally, potentially starting from templates and using standard components. After a script is developed, it can immediately become available for use or further customization.

Depending on the specific administrative function being addressed, this technology helps to tie together the various underlying components. The cells may contain numbers and text, as in most spreadsheets, as well as GUI objects such as buttons and checkboxes, visualization objects such as plots and pie charts, programming objects such as collections and timers, and system objects such as JMX, SNMP, etc. These objects can be either created by the user or assigned to cells as a result of evaluating expressions that define the functional relationship between objects in various cells.

Examples are included in the documentation provided with this package.

Contributing to this technology were engineers from several IBM locations: Almaden Research Center, Californa; Research Triangle Park (RTP), N.C.; and Autonomic Computing (AC) Division in RTP:

• Eser Kandogan, Ph.D., is a research staff member at Almaden. His interests include human interaction with complex systems, policy-based system management, ethnographic studies of system administrators, information visualization, and end-user programming.
• Eben Haber, Ph.D., works on human-comptuer interaction at Almaden. He has worked on data mining, visualization, and user interface design.
• Andrew Eberbach works for the AC team at RTP. Previously, he served as an intern in IBM's Extreme Blue program.
• Abdi Salahshour is a senior software engineer and problem determination architect for the AC group. He has served in many roles in design and development of database diagnostic tools; system management, self-healing architecture, and enablement in heterogenous and distributed environments; and several joint projects with IBM Research.
• Neeraj Joshi works as a staff software engineer in the AC group. He focuses on OSGI and the Integrated Solutions Console and has previously worked on self-healing technologies.
• Dan Jemiolo is a staff software engineer in the AC group. His main focus is the creation of developer tools that help IBM's partners and customers build autonomic systems.
• Balan Subramanian is a staff software engineer. He focuses on AC technology enablement, data collection, problem determination, and infrastructure for AC. Mr. Subramanian was a core developer on the Generic Log Adapter and Touchpoint Simulator.
• Rob Barrett, Ph.D., is a research staff member at Almaden. His current work focuses on the user experience of system administration and human aspects of autonomic computing.
• Paul Maglio, Ph.D., is a senior manager at Almaden. His group works in the areas of social, cognitive, and computer and business sciences. He holds twelve patents and has published more than 60 scientific papers.
• Allen Cypher, Ph.D., is currently working at IBM Research. His main research interests are programming by demonstration and end-user programming, in projects such as Cocoa, Eager, and "Watch What I Do."