Learning About Berkeley DB Features

The concepts behind an embedded data manager are somewhat different than other storage techniques. Berkeley DB products provide the mature, robust, scalable features found in relational products without the notion of a schema, query language, or client/server communications. Berkeley DB products are flexible customizable data management, custom off the shelf component (COTS) software, that you incorporate into your product where relational storage doesn't fit the requirements.

Oracle Berkeley DB has four primary feature sets. Each feature set is a level of functionality you, as the programmer, will choose as most appropriate for your application's needs. From basic key/data storage without transactions or concurrency to full transactional, concurrent, and replicated highly available services; Berkeley DB is a flexible customizable embedded data manager.

Each of the following tutorial presentations will cover a specific product using a specific feature set. The presentations are geared for programmers and software architects and show both source code and use case examples. These presentations are a good way to begin to understand the conceptual differences between Berkeley DB and relational databases, they will help you see how and where your product may benefit from Berkeley DB, and where you can use our embedded data manager to replace your in-house ad-hoc solution or as a component in your new development project.

• Berkeley DB Data Store (DS)

  The core features including key/data storage in hash tables, btrees, queues, and record sets. This feature set does not include support for concurrency, locking, transactions, recovery, or high availability. It is the most basic, yet fastest feature set available to developers.

• Berkeley DB Concurrent Data Store (CDS)

  This tutorial continues where the eariler Berkeley Data Store (DS) tutorial left off. It expands your knowledge of the Berkeley DB product introducing the built-in concurrency and locking features. Using Concurrent Data Store (CDS) your Berkeley DB applicaton can use multiple threads or processes to access data stored in the database. This feature set does not include support for transactions, recovery, or high availability. It builds on our most basic storage features and enables developers to build highly concurrent applications quickly and easily.

• Berkeley DB Transactional Data Store (TDS)

  Transactional Data Store adds support for ACID transactions and database recovery. This builds on the concurrent aspects of CDS providing even more functionality for developers using Berkeley DB. Using transactions groups of operations can be atomically committed or rolled back, sets of changes in one transactional context are isolated from those in others, data remains consistent within the context of a transaction, and data committed is durable and recoverable from transaction logs. This feature set is highly configurable offering developers multi-version concurrency control, dirty-read, nested transactions, long lived transactions, and other sophisticated features.
  coming soon...

coming soon...

coming soon...