The PUNCH Portal to Internet Computing: Run Any Software Anywhere - PowerPoint PPT Presentation

The PUNCH Portal to Internet Computing: Run Any Software Anywhere via WWW Browsers Nirav H. Kapadia Jos´ e A. B. Fortes Mark S. Lundstrom School of Electrical and Computer Engineering Purdue University I. Introduction to PUNCH II. System Design Overview III. External System Interface IV. Internal Architecture V. Supported Resources VI. Resource Management VII. Conclusions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

I. Introduction to PUNCH II. System Design Overview III. External System Interface IV. Internal Architecture V. Supported Resources VI. Resource Management VII. Conclusions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

PUNCH Overview: g PUNCH is a distributed network-computing infrastructure that allows geographically dispersed users to run unmodified tools via standard Web browsers ( www.ece.purdue.edu/punch ) UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Core Research Team: Nirav Kapadia, Jos´ e Fortes, Mark Lundstrom, Renato Figueiredo, and Sumalatha Adabala Collaborators: Rudolf Eigenmann, Dolors Royo, and Jos´ e Miguel-Alonso Partner Institutions: Northwestern, U. of Wisconsin at Madison, Chicago State, U. of Illinois at Urbana-Champaign, Arizona State, Stanford, U. of Texas at El Paso, SIGDA, IUPUI, and Purdue Calumet. UPC Presentation, June 2000 Nirav Kapadia, Purdue University

#Hits 800+ users 1M 3,000,000+ hits users in 10 countries usage doubles each year 1995 1999 50+ engineering tools ‘‘as is’’ installation in 30 mins operational for five years version 5 being developed UPC Presentation, June 2000 Nirav Kapadia, Purdue University

The Purdue University Network-Computing Hubs (PUNCH) The The The The . . . The Parallel Computer VLSI Semiconductor D CA ES ES RT Programming Architecture Design Simulation . . . Toolkit Hub Hub Hub Hub Network -Desktop Interface Network -Desktop Interface Front End (Network Desktop) Internet / Intranet SCION Tool (Application) Compute-Server Tool (Application) Compute-Server Software Software . . . Tool (Application) Parallel Machine Tool (Application) Parallel Machine SCION SCION . . . . . . . . Tool (Application) Workstation Cluster Tool (Application) Workstation Cluster Physical Location ‘1’ Physical Location ‘n’ UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Portal Screen Shot ( http://www.ece.purdue.edu/punch/ ) UPC Presentation, June 2000 Nirav Kapadia, Purdue University

I. Introduction to PUNCH II. System Design Overview III. External System Interface IV. Internal Architecture V. Supported Resources VI. Resource Management VII. Conclusions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Network-Computing Systems External System Interface (external view of the system) Internal Architecture (internal system design and structure) Resources Supported (class of programs/machines supported) Resource Management (application management; resource allocation) UPC Presentation, June 2000 Nirav Kapadia, Purdue University

External System Interface: g Objective: universal access g Approach: leverage the World Wide Web g Challenge: work within the existing WWW framework g Status: students and researchers in ten countries routinely use PUNCH via standard Web browsers; 50+ discipline-specific and productivity tools are available UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Internal Architecture: g Objective: scalability and interoperability g Approach: hierarchically distributed architecture g Challenge: manage tools and information in situ g Status: system deployed on Unix; support for customized user views, global user accounts, and virtual filesystems; root access not required UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Supported Resources: g Objective: work with unmodified tools g Approach: virtual interfaces; configurable environment g Challenge: state management; flow control g Status: batch, interactive, and graphical tools supported; new prototype supports PVM programs and Condor, DQS, and PBS jobs UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Resource Management: g Objective: address usage policy and performance issues g Approach: metaprograms; application management g Challenge: predictive performance modeling; interoperability g Status: scheduling customizable on a per-tool and per-user basis; prototype machine learning system facilitates automated cost/performance tradeoff decisions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

I. Introduction to PUNCH II. System Design Overview III. External System Interface IV. Internal Architecture V. Supported Resources VI. Resource Management VII. Conclusions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

External System Interface: g Problem: must work within existing WWW framework Solution: virtual URLs; web-accessible OS g Problem: HTML is static Solution: HTML templates; embedded variables/objects g Problem: no flow-control Solution: metaprograms; programmable state machine UPC Presentation, June 2000 Nirav Kapadia, Purdue University

WWW-Based Computing (Issues): WWW-Based WWW Framework Issues Internet/Intranet Computing Unique mapping between URL URL is a functional request that WWW URL model does and some entity within the is mapped in a context-sensitive not support truely dynamic filesystem (physical URL). manner (virtual URL). URL mappings. Mostly public space, with some Mostly private space, with some Authentication model for support for access control. support for sharing user data. WWW does not scale. Designed for document Primarily provides OS functions Standard WWW resource serving, with some support (e.g., process management). management model not for dynamic information. Document serving is secondary. appropriate. A read-write environment is Need synchronization and Designed for a (mostly) critical for correct operation resource locking. Support read-only environment. of OS services. for a multi-user web OS. Processing supported as an Processing is an integral part Integration of distributed add-on via CGI scripts. of the service. computing and WWW. UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Virtual Addresses: Virtual Base Offset Address Address Translation process involves a simple Translation table lookup. Physical Page Address Offset in Page Address UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Virtual URLs: . . . Virtual Base Offset Base Offset Meta-Information URL URL Translation process involves database queries, and generates side-effects (e.g., state and Translation context initialization). . . . Partially Object Handle Base Offset Meta-Information Translated URL Repeat Process for each Base-Offset Pair . . . Physical Object Handle Target Module Meta-Information ‘‘URL’’ UPC Presentation, June 2000 Nirav Kapadia, Purdue University

HTML Templates (Variables/Objects): Begin PageTemplate Begin Declarations menu myFiles = 1:<UserFiles>; bind myFiles = <WorkingFolder>; End Declarations Begin HTML <FORM> Select a file (or folder to open): <myFiles> <P> <CENTER><INPUT TYPE="submit" VALUE="Proceed"></CENTER> </FORM> End HTML End PageTemplate UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Metaprograms (Flow Control): retrievestate ’directory’; display ’Page1’; while(isdir(<myFiles>)) { chdir(<myFiles>); display ’Page1’; } savestate ’directory’; . . . UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Performance Summary: Response Time (milliseconds) Type of Transaction No Preforking With Preforking Entry Page 60 27 Tool Interface Input Management 61 27 Run Management 75 35 Output Management 61 27 Static Info Publicly-Accessible URL 47 20 Access Control Enforced 53 24 Process Status 64 30 Dynamic Information 58 29 UPC Presentation, June 2000 Nirav Kapadia, Purdue University

I. Introduction to PUNCH II. System Design Overview III. External System Interface IV. Internal Architecture V. Supported Resources VI. Resource Management VII. Conclusions UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Internal Architecture: g Problem: dynamic/incremental/distributed information Solution: hierarchically distributed architecture g Problem: multiple administrative domains Solution: resources usable/visible/invisible; partitioned root g Problem: scalability of information retrieval Solution: hashing; access codes UPC Presentation, June 2000 Nirav Kapadia, Purdue University

Core Architecture: User Client Unit User-Specific Information (e.g., tool-input, preferences, etc.) Network Interface Management Unit Tool-Specific Information (e.g., algorithmic complexity, Network platforms supported, etc.) Interface Execution Unit Site-Specific Information (e.g., default configuration, path to executable, etc.) Resources UPC Presentation, June 2000 Nirav Kapadia, Purdue University