Experiences with Performance at RIM

by Tim Barrett, RIM

Research in Motion (RIM) has just finished a 9 month project that resulted in improving the performance of Perforce by over 10x. The Project team worked with Perforce Technical Support, peer companies and internal resources to develop and execute a comprehensive project that addressed a variety of infrastructure and application issues. The team surpassed their goal of a 75% improvement in performance and has eliminated the previous performance issues. This presentation highlights the major and minor initiatives included in RIM's performance project along with key ways of measuring and understanding performance.

Demystifying Perforce Backups and Near Real-Time Replication

by Richard Baum, Perforce Software, Inc

This presentation starts with the basic building blocks of the Perforce checkpoint and journal facility. It then examines and builds on details of these, and some external tools, to show how more complex operations can be performed. Attendees will see how to implement both scheduled and near-real-time backup and hot-standby solutions. Examples will allow users to get a good feel for what is being discussed.

Perforce and Zero-Configuration Networking

by Michael Bishop, Perforce Software, Inc

The presentation will include a basic discussion of Zero-Configuration Networking, how p4d uses it to advertise servers on the network, and how clients can discover those servers. Details will include how the P4 API is affected, how developers can discover p4d in their own code, Zeroconf best-practices, and juicy miscellaneous details.

Improving Performance at Remote Sites

by Lloyd Cha & Vishnu Mohan, Montalvo Systems

Sharing of data across high-latency links remains a significant technical hurdle faced by distributed development teams. The Perforce Proxy provides an efficient and reliable way of caching data for use by developers at distant sites. However, the architecture of the Perforce system requires the central Perforce Server to track the state of all clients, both local and remote. When latency between sites is high, there is a significant cost for exchanging information with the Perforce Server. The evolution of a system to address these shortcomings will be described, and the challenges that were faced will be discussed. The development of tools to rapidly retrieve data directly from the Perforce Proxy server will be detailed.

Perforce Usage in the Games Industries

by Frank Compagner, Guerrilla Games

Developing state of the art video games has become a very data intensive operation over the last few years, and many game developers are now using Perforce to cope with this. Although games development is superficially not that different from other software development, there are some notable differences, such as a very large head revision consisting of mostly unmergeable binary data and the large number of non-technical users on a project. Specific problems caused by these different usage characteristics will be highlighted, and (partial) solutions to these problems will be discussed. Of particular interest will be performance, branching and tool development.

What's this flag? Integrating around Deleted Revisions

by Sam Stafford & Stephanie Turner, Perforce Software, Inc.

When integrations don't work as planned, it can be unclear how to proceed. This hands-on talk demonstrates how to use our favorite Perforce tools to investigate and resolve "tricky" integration scenarios involving deleted and renamed files.

Perforce Branching: Moving Fast from Theory to Practical Application

by Tom Tyler, Perforce Software, Inc.

The core themes of branching theory and the Mainline Model are fairly well established. However, there are a great many variations on the theme, as the conventional wisdom is applied to a variety of environments. This presentation helps distill the volumes of branching theory into practical recommendations for a variety of common development environments. Key factors affecting branch structure are discussed, and a set of sample case studies and resulting branch structures will be reviewed.

Large Scale Continuous Integration via Hierarchically Organized Contexts

by Thomas Kroll & Dr. Matthis Langhoff, SAP

One major challenge for SAPs Java development is the production of SAP NetWeaver, one of the world's largest Java applications. To achieve this the SAP development organization has implemented lean production principles in the last year with heavy impact on our Perforce infrastructure and operation. This presentation will give you an idea on how we tackle this challenge with Perforce.

European User Conferences: 2008 | 2006
U.S. User Conferences: 2007 | 2005 | 2003 | 2001 | 2000 | 1999 | 1998