Spring Conference 2011

Hosted by Oracle

May 19, 2011

at Oracle Corporation, Redwood Shores, CA

See directions to the conference location near the bottom of this page.

The theme for this keynote is to “Aim High.” During this presentation, Andrew Holdsworth, Senior Director, Oracle Real World Performance Group, will provide updates and insights into how Oracle Database professionals can exploit the full potential of the Oracle Database and Exadata Hardware combination.

To date, many customers have successfully implemented Oracle Database on Exadata Hardware to achieve order of magnitude performance gains when running existing applications. During this discussion, Andrew will overview the true potential of the Software/Hardware platform — when used optimally and creatively — to produce innovative solutions to old and new data processing problems.

The best practices content delivered today will hopefully inspire Oracle professionals to aspire for excellence, or “Aim High” when solving database challenges rather than applying the standard industry approach of “good enough” — because “good enough” is rarely acceptable tomorrow. This is the philosophy of Oracle’s Real World Performance Group.

It has been ten years since the publication of my original Sane SAN whitepaper and much has changed in the area of high performance database storage. In this major overhaul of the original topics of Sane SAN 2000 we look at the pros and cons of the various technologies available on the market today, and at approaches that maximise the performance of the storage tier and deliver predictable response times to the database. We also look at emerging trends in an attempt to steer new architectures in a direction that is compatible with the imminent paradigm shift in persistent storage driven by solid-state storage.

Oracle SQL Developer Data Modeler is a feature-rich product supporting logical, relational, multi-dimensional, data type and data flow models, and physical data modeling for Oracle and non-Oracle databases. In this session we follow a flow of work, starting from scratch with an empty model and touching all aspects of the product, discussing how the models relate to each other and the salient points of each. The latest release is tightly integrated with the open source, source code control software, Subversion, allowing users to track changes to models and objects. SQL Developer Data Modeler 3.0 also includes a number of other new features including the ability to create user defined design rules and transformation scripts. These and other new features will be demonstrated in this session.

This session will illustrate how PHP, together with Oracle Database 11g are being used to build mission critical enterprise and web applications. The session will start with basics of PHP connectivity to Oracle. We will then describe how customers leverage Oracle Database Resident Connection Pooling (DRCP) to handle the demanding scalability requirements of web facing PHP applications. The session will also touch upon key facilities in PHP that Oracle provides for HA, online application upgrades, result caching, and end-to-end tracing.

This beginner-level session will introduce Oracle's offerings for .NET programmers, including Oracle Data Provider for .NET (ODP.NET), Oracle Developer Tools for Visual Studio, Oracle Providers for ASP.NET, and .NET stored procedures. Step-by step-demos will be used to illustrate how to get started developing Oracle Database .NET applications using each of these free products. New and upcoming features will also be described briefly in this session, including Entity Framework, LINQ, WCF Data Services, 100% managed Thin ODP.NET, and support for TimesTen.

Large, mission-critical applications built on Oracle Database are often unavailable for tens of hours while the application’s database objects are patched or upgraded. Oracle Database 11g Release 2 introduces revolutionary new capabilities that allow online application upgrade with uninterrupted availability of the application. Existing sessions can continue to use the pre-upgrade application until their users decide to finish; and, at the same time, new sessions can use the post-upgrade application. When no sessions are any longer using the pre-upgrade application, it can be retired. The application as a whole therefore enjoys hot rollover from the pre-upgrade version to the post-upgrade version.

The capability depends on these new kinds of object: the edition, the editioning view, and the crossedition trigger. Code changes are installed in the privacy of a new edition. Data changes are made safely by writing only to new columns or new tables not seen by the old edition. An editioning view exposes a different projection of a table into each edition to allow each to see just its own columns. A crossedition trigger propagates data changes made by the old edition into the new edition’s columns, or (in hot-rollover) vice-versa. The capability as a whole is called edition-based redefinition – EBR for short.

This session explains how it all works.

Attend this session to learn best practices for deploying a data warehouse on Oracle Database 11g. Using detailed examples, this session will cover the best practices for schema design, hardware configuration, data loading, partitioning strategies, parallel query, and workload management. You will also learn how to monitor, diagnose, and correct common performance issues using both the old reliable tools such as explain plan as well as an introduction to the latest and greatest SQL monitor tool. By applying what you learn in this session you will be able to deploy a data warehouse that can seamlessly scale without constant tuning or tweaking.

The Exadata Database Machine offers new levels of processing performance and a new set of concepts to learn. Based on my book review of “Achieving Extreme Performance with Oracle Exadata” (Oracle Press), this presentation covers what is new and what is familiar to an Oracle DBA. Due to the combination of hardware and software that make up the Exadata machine, there are many new things to learn including the storage cells, a new command line interface and the requirements a database must meet before it can be migrated to Exadata.

Replication seems like the silver bullet to solve scalability, reliability and geographic issues. Oracle has deprecated Streams and acquired Goldengate as the preferred solution for hot-hot multimaster replication. However there are systemic considerations for multimaster systems in general and the current state of the art for Goldengate in particular that need to be considered from an application perspective before committing to this replication technology. To replicate DDL or not, partitioning of sequences and data domains geographically and the impact of replication lags especially with RAC are some of the issues explored in this talk. Learn which systems this powerful tool is best applied to and when the operational complexities make it problematic.

In this session we will cover the basics of RMAN. We will look at the architecture of RMAN and basic backup and recovery guides. We will also look at the use of an RMAN catalog. In closing we will mention a few tips and tricks when using RMAN. At the end of the session you should have a basic understanding of RMAN and be able to do basic backup and recovery with your Oracle databases.

These presentations explore and troubleshoot performance issues encountered in RAC environments. Below are a few discussion points:

• Trend analysis for global cache traffic and performance using AWR/statspack tables. Review of simple techniques to probe interconnect performance.
• Understanding wait events in SQL trace files related to RAC such as gc cr grant 2-way, gc cr grant 3-way, gc multi block reads, etc.
• Global cache object re-mastering
• Global cache waits and LGWR performance. LGWR performance is important in single instance databases and LGWR performance is even more critical in RAC instances.
• Interconnect issues, lost packets, and statistics. Identifying interconnect issues with database and OS statistics.
• Troubleshooting performance issues with MTU/jumbo frames, and jumbo frames mis-configuration issues.
• Troubleshooting RAC background processes such as LMS/LMD, etc.
• Troubleshooting parallel query, interconnect, and PQ setup related performance issues.
• Concurrency-induced global cache issues and a few time-tested simple work-arounds.
• Troubleshooting locking issues and deadlocks in RAC environments.
• Bind peeking issues and RAC instances.
• ...and a few misc items.

Most presentations on SQL performance assume that the query developer is highly experienced in SQL and Oracle and focus on tracing and other technique. Sadly, in today's world, knowledge of SQL and Oracle are not highly valued, and many SQL performance problems are the result of inexperience, misguidance, “heard it/read it somewhere” and various source of mis-information. This presentation will address common “anti-patterns” in database design and SQL query writing that I've encountered over the years and how to fix them without tracing, changing init.ora parameters or techniques requiring DBA level privileges. It is focused on techniques that are typically available to developers.