The Progress Enterprise RDBMS

The Progress Enterprise RDBMS is designed for large user environments and the transaction processing throughput of today's most demanding SQL- and Progress 4GL-based on-line transaction processing (OLTP) applications. Grounded in a flexible, multithreaded, multiserver architecture, the Enterprise RDBMS is a powerful, open and large-scale enterprise database that can run across multiple hardware platforms and networks.

The Enterprise RDBMS includes all of the functionality needed to meet the most demanding OLTP requirements. These capabilities include row-level locking, roll-back and roll-forward recovery, point-in-time recovery, distributed database management with two-phase commit, a complete suite of on-line utilities and full support for ANSI-standard SQL-92. The unique combination of power, flexibility and ease of operation makes the Enterprise RDBMS an ideal engine for a wide range of commercial and data processing applications.

Sophisticated self-tuning capabilities and simple graphical interfaces for system administration make the Enterprise RDBMS easier to install, tune and manage than other products. With low administration costs, low initial cost of licenses, minimum upgrade fees and limited software implementation costs, the Enterprise RDBMS provides a significant cost-of-ownership advantage - approximately four to one - over competing database products.

Supports Multiple Operating Systems and Networks

To accommodate the requirements of complex enterprise networks, the Enterprise RDBMS runs on a wide range of hardware platforms and operating systems, from PCs running Windows or Windows NT® to large multiprocessor systems. Benchmarks show proven scalability from 1 to 32 processors in both SMP and ccNUMA architectures.

Grows with You

A flexible architecture and performance optimization features allow the Enterprise RDBMS to scale according to your business requirements. You are limited only by your system resources. The Enterprise RDBMS provides:

 

	Support for up to 10,000 concurrent users and for databases in excess of 80 million terabytes. These features, combined with support for 64-bit operating systems, provide exceptional capacity for large-scale, high-performance computing.
	Support for up to 32-way SMP and ccNUMA configurations. Fine-grained shared memory locking and SMP spin locks provide highly concurrent processing for on-line interactive users by reducing contention for shared memory. The resulting performance advantages allow the Enterprise RDBMS to scale effectively from smaller servers to the largest enterprise servers.
	A large buffer pool that allows an increased portion of the database to be resident in memory. Support for up to 125 million buffers, or 128 Gigabytes, in a 64-bit OS environment provides memory-speed access to the frequently used parts of the database and ultimately reduces disk I/O.
	A high-performance, multithreaded, multi-server architecture. This unique architecture provides efficient use of system resources and maximizes concurrency while minimizing network utilization.
	Row-level locking to reduce resource contention and maximize performance. This important feature enables efficient, safe, data sharing in multi-user systems by eliminating locking of an entire block or page of information when only a single record is in use.
	Optimistic transaction logging. By logging only recoverable transactions, the Enterprise RDBMS enables smaller roll-back logs and improved reuse of log space.
	Self-tuning server processes. Self-tuning features, such as Asynchronous Page Writers, adjust their behavior based on database activity. These features optimize performance even in instances, such as switching from on-line processing to batch processing, that involve dramatic change.
	Dynamic buffer pool allocation for batch and interactive users. Allowing a batch process or online user to isolate table scans to a small portion of the buffer pool allows large scan and reporting tasks to have almost no effect on the database buffer pool. This in turn greatly reduces the negative impact on the system from large scan operation.

Protects Your Data and Provides High Availability

The Enterprise RDBMS provides all of the security, data protection, and high availability capabilities required by mission-critical applications. Features such as backup (on-line and off-line) and restore, two-phase commit, roll-back, roll-forward and point-in-time recovery capabilities and automatic crash recovery, form an industrial-strength armor to protect your data.

The Enterprise RDBMS provides:

 

	Automatic crash recovery. The Enterprise RDBMS performs crash recovery automatically when the database is restarted following a failure. Recovery is efficient and effortless.
	Roll-back logging. This feature provides protection for the database in the event of a database failure. The database uses the roll-back log during crash recovery to apply committed changes to the database and roll back all non-committed transactions.
	Roll-forward logging. The roll-forward feature, which replays committed transactions from the point of backup, enables point-in-time recovery and cluster failover support. Failover for both "shared-disk" and "shared-nothing" configurations is supported.
	On-line full and incremental backup. To eliminate the need for scheduled downtime for backup, the Enterprise RDBMS supports on-line backup during normal processing. The high availability afforded by this feature supports "non-stop" applications.
	On-line table and index reorganization. The Enterprise RDBMS provides the ability to reorganize a table while the system is in operation and users are operating against the table with no write locks. Users can continue forward processing with no restrictions while indexes are being rebuilt.
	Two-phase commit. The Enterprise RDBMS implements a two-phase commit protocol as an automatic feature to ensure the integrity and consistency of transactions that span multiple databases and multiple sites. This protocol ensures that a transaction can be rolled back if a failure occurs at any node the network.
	Point-in-time recovery. The database can be rolled forward from a backup to a particular point in time. This allows recovery from an inadvertent user error or some other non-system type of failure.

An Open Solution

The Enterprise RDBMS meets the ANSI SQL-92 Entry-Level standard, providing an open environment that enables efficient integration with third-party tools, such as development software, reporting and OLAP tools.

Featuring a state-of-the-art, cost-based query optimizer, the Enterprise RDBMS SQL engine is a modern, high-performance data processor. It incorporates SQL client access APIs for ODBC, JBDC, and Embedded SQL/C, and provides SQL extensions for a more robust solution.

Optimized transaction logging minimizes resource contention and space usage. The database logs only those operations that need to be fully recoverable. For example, when a user connects to the database to run a report, nothing is logged to the roll-back log since the user never issues a recoverable operation, such as an insert or an update.

Additionally, by re-writing inefficient queries prior to execution via a Predicate Re-write feature, the SQL engine improves performance and simplifies developers' understanding of proper SQL query construction.

Additional SQL extensions include:

 

	Java stored Procedures and Triggers
	GRANT/REVOKE security model
	Complete SQL DDL implementation including ALTER TABLE
	Updateable views
	Derived Tables
	Multi-Schema support

Simplifies Administration

In addition to self-tuning features, the Enterprise DBMS incorporates the Progress Explorer administration tool, which manages Progress databases and servers across the network, simplifying administration. Progress Explorer supports command-line scripting, allowing a database administrator to store scripts in a single location and use them to start/stop and manage systems simultaneously across multiple platforms.

Supports Distributed Database Access

The Enterprise RDBMS provides easy, complete access to data stored in multiple databases across a network. Applications can connect to as many as 240 databases simultaneously.

 

	Highly scalable solution featuring optimized performance
	Up to 10,000 concurrent users and large buffers
	Multithreaded server; multiple servers per database (up to 256)
	Support for multiple operating systems and hardware platforms
	ANSI SQL-92 Entry-Level compliance plus extensions
	Built-in SQL client access APIs
	Comprehensive reliability features, including automatic crash recovery, commit/roll-back transactions and roll-forward capabilities
	On-line backup, cluster fail over support, and table and index reorganization for high availability
	Dynamic user-level buffer allocation to prevent table scan operations from affecting the overall buffer pool
	Virtually unlimited data storage, limited only by hardware

The Progress Workgroup RDBMS

The Progress Workgroup RDBMS, which offers many of the same powerful capabilities as the Enterprise RDBMS, is optimized for workgroups of 2 to 30 simultaneous users. This cost-effective, department-level solution provides high performance, multi-user support, and cross-platform interoperability - at an excellent value. It meets the needs of workgroup applications by running on a wide variety of hardware and operating system platforms. Because the flexible database architecture provides optimal throughput on all platforms, a database developed on one machine can serve applications on other systems and network configurations.

As organizations grow larger than 30 users, they need a simple license upgrade to go from the Workgroup to the Enterprise database. Not a single line of application code must be changed.

The Progress Personal RDBMS

Bundled with Progress development tools, including Progress ProVision Plus, the Progress Personal RDBMS is the optimal choice for deploying single-user SQL- and 4GL-based applications and for developing, prototyping and testing applications.

 

Database Parameters Maximum:	Enterprise RDBMS	Workgroup RDBMS	Personal RDBMS
Tables	32,000	32,000	32,000
Indexes	32,000	32,000	32,000
Fields per table	32,000	32,000	32,000
Concurrent users   per database	Up to10,000	Up to 30	1 per database
Record length	32,000 bytes	32,000 bytes	32,000 bytes
Buffer pool	125,000,000 buffers/ 128 Gigabyte (64-bit), 500,000 buffers/ 4 Gigabyte (32-bit)	500,000 buffers/4 Gigabyte (32 bit)	500,000 buffers/4 Gigabyte (32 bit)

Data Types Supported:

 

Character	32,000 bytes (maximum)
Date	1/1/32768 BC to 1/1/32767 AD
Decimal	50 digits total, 1-10 decimal places
Integer	-2,167,483,648 to 2,167,483,648
Logical	true/false, yes/no

 

Network Support: TCP/IP

	Platforms Supported: Windows® 95/98 Windows NT®/Intel® Compaq Tru64 UNIX DG/UX Intel HP-UX® IBM AIX® SCO OpenServer™ SCO UnixWare™ Sun® Solaris™ SPARC Sun® Solaris™ Intel Citrix® MetaFrame
	AIP Support: ANSI SQL-92 Progress 4GL ODBC 3.0 JBDC Embedded SQL/C TCP/IP