This version includes improved support for the new alternative query mechanism. This new functionality originally introduced in version 3.40 provides application developers with a means of defining alternative SQL queries to override the default queries generated by the DBPronto server.

Alternative queries can be more efficient and potentially faster than those generated by DBPronto since they can take advantage of the application developers in-depth knowledge of their system.

Alternative queries can include joins that enable multiple tables to be read using a single query. When using such joins care must be taken, as if too complex they may have a detrimental affect on performance.

The alternative query mechanism requires only minimal program changes since the queries and how they relate to I/O statements are defined in metadata files. The mechanism currently only supports the READ and REWRITE statements on COBOL index files.

Please refer to the DBPronto documentation for more information.

Have fun.

DBPronto server Table Caching

DBPronto can be made to use table caches for improving the performance of table read operations, especially when performing random read operations. The DBPronto server will automatically update a table cache when performing updating operations such as OPEN OUTPUT, WRITE, REWRITE and DELETE.

Please refer to the exciting, newly overhauled DBPronto documentation for more information.

In the near future we plan to enhance this functionality so that it can be used with the client block reading mechanism so as to reduce the DBPronto client/server communication and provide further performance benefit.

Identifying owners of locks from the DBPronto Administrator

The DBPronto Administrator has been enhanced to display the PID and user name of users connected to the DBPronto server. With the session ID and the list of record locks, this information may be used to identify the users who have particular records locked.

Opening over 250 files

The DBPronto client libraries have been enhanced so that the maximum number of file’s that can be opened at one time by a process is now configurable. Previously this value was limited to 250.

New API calls to fetch multiple rows using an SQL query

The problem with the DBPronto SQLCMD API routine is that when used with SQL queries, it may only be used with those that return a single row. To overcome this restriction we have introduced the SQLQUERY, SQLQUERY_FETCH and SQLQUERY_CLOSE API routines.

This functionality may prove useful in the few cases where existing program logic doesn’t work well with DBPronto and standard file I/O, and can better be expressed with alternative SQL coding. An example of this are batch jobs where large volumes of data are read but performance is critical.

In the near future we plan to enhance this functionality so that it can be used with the client block reading mechanism .

Table Maintenance routines

We have enhanced the DBPronto client libraries with new routines to enable application programs to instruct the DBPronto server to globally close and reopen tables, and perform metadata refresh operations.

The close and reopen routines may be used to prevent applications from accessing tables whilst for any reason they have to be taken off-line. The reopen routine may also be used to create tables that do not already exist.

Using SET_WHERE with the DBPPP

Enhanced the DBPronto pre-processor so that it will recognise a new DBP-SET-WHERE statement that will be converted to a call to the DBPPP_SET_WHERE routine. This enhancement was especially useful for RM COBOL application that wish to use the SET_WHERE functionality which may be used to restrict rows selected by queries.

The metadata creator to sort alternate keys by byte order instead of SELECT order.

We have enhanced the metadata creator to define alternate keys by byte order instead of SELECT statement order. This behaviour is now the default when generating metadata for use with ACU but can also be enabled and disabled using specific command option lines options.

One of the big enhancements is the implementation of configurable table caching.

To give you an indication of the benefits, the table below shows the time taken by a program using U/FOS data access to read 8000 records from a  U/FOS database compared with the same program reading from a SQL Server database via DBPronto with a variety of configuration options:

This demonstrates the fact that DBPronto performance can be improved using a range of configuration options which are designed to be applicable in different situations depending on the data structures used and the type of processing.

Another option is to use DBSQLCMD and DBFETCH commands to access the data using specific SQL. This can be very helpful in the very few cases where existing program logic doesn’t map well onto RDBMS data access and the problem is difficult to solve using DBPronto configuration alone.

It is now much easier to navigate and includes a new, easy-to-use ‘Getting Started’ section.

Please download DBPronto.chm from ftp://ftp2.transoft.com/pub/DBPextras/

• Windows 2000, Server 2003, 2008, Windows XP, Vista, 7
• Linux, PowerPC or x86, 32 or 64 bit
• HP-UX PA-RISC
• IBM AIX 64 bit
• Solaris SPARC
• Solaris x86
• SCO OpenServer 6

For full update details, including bugfix details, please refer to the DBPronto release notes, available here:

ftp://ftp2.transoft.com/pub/DBPextras/DBPRelease.htm

Additional functionality since v2.50

• Solaris/x86 is now a supported DBPronto platform.
• SCO OpenServer 6 is now a supported DBPronto platform. Most DBPronto client components on SCO OpenServer6 are built with the OSR5 ABI to allow linking with AcuCOBOL runtime etc, (except for “libdbpISAM”, which is built to the SVR5 ABI to provide support for multithreaded C applications). All server interface components are built with the SVR5 ABI, (to allow access from the DBPronto server via JNI [Java Native Interface]).
• DBPronto server now reports errors when loading the metadata of files whose record or key lengths exceed the configured MAX_RECORD_LENGTH and MAX_KEY_LENGTH settings.
• Added a DBPronto server configuration option USE_CASCADING_QUERIES that will cause cascading queries be used on all files. (Previously this could only be set in metadata. This can still be overridden in metadata).
• Enhanced DBPronto server to automatically disable block reading (caching) when it finds that cached records are being discarded. The mode will be automatically be re-enabled when the DBPronto server finds that sequential reading has been resumed.
• Views can now be specified in the metadata to affectively join multiple tables. The ‘where’ attribute used in view definitions can specify fairly complex pseudo-SQL.
• Added I/O Intercept routines dbpEXTFH_is_managed, dbpEXTFH_get_locked_sessionID and dbpEXTFH_get_user_info to get lock information.
• I/O Intercept now supports DBWHERE.
• DBAPI (libdbp/dbp_client) now supports DBP_GET_USER_INFO
• DBPPP (RMCOBOL) now supports data records of more than 9999 bytes.

Additional functionality since v2.30 / v2.40

Note: v2.40 was not publicly released.

• Enhanced the DBPronto server so that refreshing the metadata from the DBPronto Server Administrator will create any subtables that have been introduced. If a subtable already exists in the RDBMS a warning message will be reported and the old subtable will kept.
• Changed the DBPronto server to report a file status of 30 instead of 23 or 10 when a serious error is encountered when processing a query for a READ request.
• Enhanced the DBPronto server to enable the SQL Server 2008 DATETIME2 data type to be used. This requires SQL Server 2008 and the Microsoft JDBC 3.0 driver, (JTDS does not support the DATETIME2 data type as this time, (JTDS feature request 3014595)). Metadata attribute ‘sqltype=”datetime2″‘ added to support this. Note that the Microsoft JDBC 3.0 driver connection  string is different from JTDS connection string. (See DBPronto documentation for more details).
• Enhanced DBPronto server to perform reconnection attempts on connection reset when in DBAM mode. (Previously, with DBPronto in DBAM mode, no database reconnection attempts were made).

• Windows Server 2000, 2003, 2008, Windows XP, VISTA, 7
• Linux, PowerPC or Intel, 32 or 64 bit
• HP-UX PA-RISC
• IBM AIX 64 bit
• Solaris SPARC

A large number of features are included in this release and other product releases since v2.0.

Additional Functionality

• Support for Fujitsu COBOL inthe DBPronto I/O Intercept libraries.
• Transaction processing is now supported in the C-ISAM/D-ISAM client libraries.
• Add support for time and timestamp field types
• Allow definition of return value to be used when reading sanitized date and time fields
• Enable configuration of C-ISAM MAX_REC_LEN value.
• It is now possible to load changes to the metadata or load new metadata without stopping and starting the DBPronto Server. A ‘Refresh Metadata’ button has been added to the DBPronto Administrator.
• Use KEEPALIVE to ensure the DBPronto Server is notified when client processes disconnect. This ensures resources such as locks can be released.
• Improvements in the .NET API to the Lock Server
• If the DBPronto Server connection to the RDBMS is broken, allow configuration for re-connect.

Performance Improvements

• When using DBPronto I/O intercept client libraries and passing file I-O to the native COBOL file handler
• Use stored procedures for READ operations when keys are composed of multiple fields; also for DBAM READ and WRITE.
• Allow the use of joined sub-queries for DBAM
• The DBPronto Server and Client can now be configured to communicate via named pipes rather than sockets.

DBPronto Pre-Processor and DBP API client library

• Provide emulation of RM-COBOL C$RERR
• Provide emulation of RM-COBOL specific START FIRST/LAST syntax.
• Ensure the pre-processor handles  RM-COBOL clauses: SELECT NOT OPTIONAL, NOT INVALID, NOT AT END
• Implement a pre-processor to convert ACUCOBOL Transaction statements to DBP API calls.

Metadata Creation and Management

• Allow definition of a condition for a sub-table to avoid empty rows in the sub-table.
• Create all definitions for U/FOS DBAM files
• The metadata creator has been enhanced to allow definition of the  record prefix, a file containing ‘replace’ words, and an option specifying the redefine to use.

The DBPronto Schema Converter converts an SQL schema into a COBOL data format which can be used by COBOL programs to access the database, and also into an XML format which can be accepted by the DBPronto Server to control access the RDBMS.

The converter accepts a list of SQL table names; for each table it uses JDBC metadata commands to read the table metadata, which is then processed to produce SELECT and FD statements for use by a COBOL program, and also an XML file for use by the DBPronto Server.

The schema converter will attempt to produce output which will be acceptable to both a COBOL compiler and also the DBPronto server. However, this is not a process that can be guaranteed to succeed; for example, an SQL field name might clash with a reserved word for a particular compiler, or a field size might not be acceptable. The schema converter provides some options that can help in this case, but it is still possible that manual editing of the output files will be needed.

Click here to download the DBPronto Schema Converter.

• Windows Server 2000, 2003, 2008, Windows XP, VISTA, 7
• Linux PowerPC or Intel, 32 or 64 bit
• HP-UX PA-RISC
• Solaris SPARC and Intel

The major features of this release are the addition of support for U/FOS DBAM (Data General INFOS) file access as well as support for C-ISAM/D-ISAM file access.

Applications that previously accessed U/FOS or C-ISAM data can now access the data migrated to RDBMS tables with no changes to existing source code.

The new features in DBPronto 2.0 include:

Support for DB2. The databases that can be accessed by the DBPronto Server are SQL Server, Oracle, MySQL, PostgreSQL, Sybase, and now, DB2.

Support for C-ISAM/D-ISAM file access. Applications that access C-ISAM/D-ISAM data can now access RDBMS data with no code changes, it is not even necessary to re-compile.

Support for U/FOS DBAM file access. U/FOS, originating from Data General INFOS,  is a complex hierarchical database management system. Applications that use U/FOS can now access RDBMS tables instead without making any changes to the application code. DBPronto fully implements the functionality provided by the U/FOS file access commands.

DBWHERE performance feature. DBPronto has been enhanced to enable application programs to specify a partial WHERE expression that will be included in the query executed for a subsequent READ NEXT/READ PREVIOUS operation.

Flexible filenames. The VARFN feature has been extended so that filenames can be specified using any regular expression.

CBL_CHECK_FILE_EXIST. The DBPronto EXTFH client library has been enhanced to provide support for the CBL_CHECK_FILE_EXIST runtime routine.

DBP_GET_USER_INFO. The DBPronto client API library has been extended to provide this new command that returns information related to the session ID, this can be used to identify the owners of locked rows for example.

CHANNEL_LOCKING_MODE. This is a new DBPronto Server configuration option that enables definition of the way in which locking works when a file is accessed by the same client on multiple channels.

For more information, please download the DBPronto 2.0 Release Notice

• Windows Server 2000, 2003, 2008, Windows XP, VISTA, 7
• Linux PowerPC or Intel, 32 or 64 bit
• HP-UX PA-RISC
• Solaris SPARC

New features in the DBPronto 1.4 release include:

Block Read Mode in the DBPronto Server. This provides performance improvements where the client application is reading sequentially through the data.

PostgreSQL performance has been improved by modifying the query syntax to ensure the query plan is optimal.

Support for Micro Focus COBOL Filename Mapping.

If environment variables are set in one of the following formats, DBPronto will translate them at “open” time:

• dd_FILENAME=DESTNAME
• DD_FILENAME=DESTNAME
• FILENAME=DESTNAME

Client Block Read

Block Read Mode can be used to improve performance by reducing the network traffic between the RDBMS and the DBPronto Server. Block Read Mode and Client Block Read Mode both improve performance in scenarios where the client application is reading a large number of records using READ NEXT or READ PREVIOUS.

If Block Read Mode is configured, Client Block Read Mode can also be used to reduce the network traffic between the DBPronto Server and the DBPronto client application.

The Metadata Creator will now ignore any records which have no indexes if command line option, x is specified.

Transoft, an expert in innovative modernization software and service solutions, today announced the launch of DBPronto. Designed to work with legacy file handlers including COBOL, C-ISAM and Transoft U/FOS, DBPronto provides a rapid, easy and automated means of allowing legacy applications to use a modern relational database (RDBMS) such as Oracle, MySQL, PostgreSQL and Microsoft SQL Server – replacing older proprietary ISAM file handlers.

With Transoft DBPronto you can:

• Improve data integrity and robustness
• Implement database high availability solutions
• Readily share data with other applications
• Easily implement data warehousing and business intelligence solutions

DBPronto is based on Transoft CIO which was launched 6 months ago and is already being used successfully in a number of organizations. DBPronto extends the functionality to a wider range of legacy data structures, and provides a bridge between existing code and the new RDBMS.

“Transoft’s revolutionary new DBPronto product has enabled us to combine our existing Insurance Broker application with our new .NET based Life Insurance system without having to make extensive changes to either. This has opened up new markets for us with existing customers, as well as enabling us to provide a single integrated system.”
Alastair Bell, Managing Director, Relay Software