Data Profiler 4.2 on Linux - Readme
General Release - February 2008

Contents

Welcome to Data Profiler release 4.2. This document contains the most up-to-date information for running this product on Linux systems and may supersede that in the product documentation. It covers the following topics:

System Requirements

The following hardware and software is needed for a minimum and preferred installation:

Installation Utility

There should be no differences in installing Data Profiler on either the Red Hat or SUSE Linux operating systems. The installation creates a Groupid called pvsw, under which dataprofiler.sh runs. The pvsw Groupid permissions are 644. Profiles and the resulting files run by a user are owned by the user who created them.

Preinstallation Checklist

Review the following checklist before installing Data Profiler:

To Install from a CD

  1. Log on to the machine as a root user.
  2. Mount the CD.

    3. cd /<media>/<cdrom>

    For media and cdrom, use your CD device name.

  3. Create a temporary directory. This example uses mytmp:

    4. mkdir /mytmp

  4. Copy the archive to the directory:

    5. cp DataProfiler-<version>-<date>.tar.gz /mytmp

    For version and date, use the values from the .tar.gz file name.

  5. Change directories to the temporary directory:

    6. cd /mytmp

  6. Uncompress the archive:

    7. gzip -d DataProfiler-<version>-<date>.tar.gz

  7. Untar the archive:

    8. tar -xvf DataProfiler-<version>-<date>.tar

  8. Run the installation utility:

    9. sh dataProfiler-<version>-<date>.sh

  9. Read the license agreement and type Yes to accept. The software installs.

What to Do Next

After installing, you must apply the license file. See To Apply the License File.

To Install from a Downloaded Archive

  1. Log on to the machine as a root user.
  2. Create a temporary directory. This example uses mytmp:

    3. mkdir /mytmp

  3. Copy the archive to the directory:

    4. cp DataProfiler-<version>-<date>.tar.gz /mytmp

    For version and date, use the value from the .tar.gz file name.

  4. Change directories to the temporary directory:

    5. cd /mytmp

  5. Uncompress the archive:

    6. gzip -d DataProfiler-<version>-<date>.tar.gz

  6. Untar the archive:

    7. tar -xvf DataProfiler-<version>-<date>.tar

  7. Run the installation utility:

    8. sh dataProfiler-<version>-<date>.sh

  8. Read the license agreement and type Yes to accept. The software installs.

What to Do Next

After installing, apply the license. See To Apply the License File.

To Apply the License File

The following are instructions for applying the default license file. If you do not have a license file, see Obtain a license file.

  1. After you have installed Data Profiler, copy the file dataprofiler.slc to the following location:

    2. /opt/PervasiveSoftware/DataProfiler<version>/com.pervasive.profile.engine/license

  2. Change the file access permissions as follows:

    3. chmod 644 /opt/PervasiveSoftware/DataProfiler<version>/com.pervasive.profile.engine/license /dataprofiler.slc

What to Do Next

After installing, you may want to verify the installation. See Postinstallation Steps.

Postinstallation Steps

To ensure that Data Profiler is set up correctly, follow these steps after installation:

  1. Set the appropriate environment variables for all Data Profiler users:

    2. export PATH=$PATH;/opt/PervasiveSoftware/DataProfiler<version>/com.pervasive.profiler.engine/bin

  2. Run the following command to verify the installation:

    3. rpm -qi -a | grep DataProfiler

Reverting to Previous Unix Permissions

If the security settings applied during installation are not appropriate for your environment, you can revert to your previous permissions by executing the following commands as root:

chmod -R 755 /opt/PervasiveSoftware/DataProfiler<version>/dataprofiler

chmod -R 777 /opt/PervasiveSoftware/DataProfiler<version>/plugins

chmod -R 777 /opt/PervasiveSoftware/DataProfiler<version>/jars

chmod -R 777 /opt/PervasiveSoftware/DataProfiler<version>/license

The Unix security feature for Data Profiler is now disabled.

Exporting Projects to .Jar Files

To prepare Data Profiler projects to run on the command line, you must export each project to a .jar file. This is done in the Data Profiler user interface on Windows.

To export project files

  1. Open Data Profiler.
  2. In the Project Navigator pane, right-click and select Export.
  3. Click the plus sign next to Data Profiler and select Export Data Profiler project. Click Next.
  4. In the left pane, select the project name you want to export.
  5. Next, to select an export destination to run a profile on a remote machine, select Create a .jar archive file.
  6. At File, type the name of a new file to create, or click Browse to navigate to an existing folder and then append the file name at the end of the path. The default location for the export operation is <Drive>Documents and Settings\All Users\Application Data\Pervasive\Data Profiler 4.2\workspace.
  7. Click Finish.

What to Do Next

After creating a .jar file, you are ready to run a profile on the command line. See Running Profile Specification Files on the Command Line.

Running Profile Specification Files on the Command Line

After you have finished designing profile specification files in the user interface, you may want to run those .dp.xml files on the command line.

Prerequisites

Before you run profiles on the command line, you must have done the following:

To run a .dp.xml file from the command line

  1. Open a command prompt window. On the command line, type the command to call the profile.

    2. Example:

    dataprofiler.sh -cp projects/local/clients/sales.jar -s Profiles/topsales.dp.xml

    The project classpath (-cp) and the profile name (-s) are required. The project classpath contains the profile to run and all the resources that the project depends upon, such as data set descriptions. The .jar file is a resource in the project classpath, so its location should be specified using a project-relative path with forward slashes. The .jar contains the source file, connection or data set, profile file, schema file, and project file.

  2. View the console output. It should look similar to the following:

    3. INFO, 2008-01-24T20:13:03.697Z, Preparing profiler engine

    INFO, 2008-01-24T20:13:03.822Z, Profiler engine is ready to run jobs

    INFO, 2008-01-24T20:13:03.837Z, Preparing profiling specification Profiles/topsales.dp.xml

    INFO, 2008-01-24T20:13:05.868Z, Executing profiling specification Profiles/topsales.dp.xml

    INFO, 2008-01-24T20:13:05.947Z, Finished profiling specification Profiles/topsales.dp.xml

    INFO, 2008-01-24T20:13:05.947Z, Specification run time: 00:00:03

    Output files are created in a Results folder child directory of the current directory.

Command Line Options

Supported command line options are listed in the following table.

Table 3-10 Command Line Parameter Options
Command
Description
-cp
Required. Project classpath. Load dependent specifications from directories in this classpath (separated by semicolons).
The classpath must include the root directory of your project or of the .jar file to which you exported your project. This resolves resource paths relative to the project root.
-h
Print command line help and exit.
-l
Write log messages to a file. Include path and name of the log file for the profile. If none is given, the current working directory is assumed.
-m
Memory heap size.
  • Append M for megabytes. For example, -m 800M
  • Append G for gigabytes. For example, -m 1G
Best practice suggestion: If you have performance issues and you have the system resources, begin using the default of 512M. If performance is still an issue, try a greater value. In addition, you can also pass additional JVM tuning properties. For details, see Tuning the Data Profiler Engine for Improved Performance.
-r
Write results to this directory. Include path and name of the results folder that contains profile output files. If none is given, current working directory and the default Results folder are used.
-mpf
Specify the macro property file names. Provide the full path to each file, except in the following cases:
  • Your macro definition file is in the classpath.
  • You pass a .jar file that resides in the class path and the .jar file contains the macrodef.xml file.
You can specify multiple macro definition files. The last macrodef.xml file listed on the command line takes precedence over the other files.
-mp
Provide a value for a macro in name and value pairs. You can override previous values with this command. This option takes precedence over the -mpf option. This flag is used for individual macro settings, and more than one property can be passed at a time. If you provide more than one value, the last value takes precedence.
-rt
Specify an output report type. Options are csv, html, and pdf
If you do not specify a format, reports are not generated.
-s
Required. Run the profiling specification. Include path and name of the profile specification file. If no path is provided, the current working directory is assumed.
-version
Provides the version of Data Profiler application.

Example With Optional Parameters

/opt/PervasiveSoftware/DataProfiler<version>/com.pervasive.profiler.engine/bin/dataprofiler.sh -cp projects/local/clients/sales.jar -mp MYMACRO=/datafiles/dp42/projects -m 256m -l ./log_files -r /datafiles/dp42/projects/results -rt csv -s Profiles/topsales.dp.xml

This example specifies the path to the .sh file, path to the project .jar file, macro name and value, memory heap size, log file folder name, results folder name, output report type, and path to the profile file.

Tuning the Data Profiler Engine for Improved Performance

You can tune the data profiler engine Java Virtual Machine (JVM) to improve performance when running profiles and to address resource exhaustion issues, such as OutOfMemory errors. JVM tuning controls the following factors:

When you run the engine from the command line, the -m flag sets heap memory size, but you can tune the engine further.

The dataprofiler.tuning.sh file included in your product installation provides additional JVM arguments that you can set. The file is located in the following directory:

/opt/PervasiveSoftware/DataProfiler<version>/com.pervasive.profiler.engine/bin

Open the file in a text editor to read the JVM argument notes and to see the options. You may uncomment and tune the code in the file as needed.

See Also

Running Profile Specification Files on the Command Line

Known Issues and Resolutions

This section lists the most noteworthy known issues for Linux as of the release date. If you encounter an issue not found here, contact Pervasive Support.

Issue
Resolution
You do not have administrative privileges for the machine. A root user ID and password are required to install the product.
Contact your support organization and request that your privileges be modified to Administrative for the affected machine.
While performing a maintenance installation, a message is returned saying your subscription is expired.
Contact your support organization to obtain an updated license. After implementing the new license, attempt to install again.
You have trouble running Data Profiler and encounter a loading of shared libraries error.
Put the Data Profiler installation directory at the beginning of your PATH environment variable. For details, see Postinstallation Steps. In rare cases, customers have reported problems with an existing incompatible version of one of the third-party components that we use, such as ACE, ICU, or Xerces C. Place Data Profiler at the front of the search paths to insure the component version in the Data Profiler install directory is loaded first at run time.
34877 - A profile that includes an IBM DB2 7.2 connection fails and returns an error.
Run any profile that uses an IBM DB2 7.2 connection on a supported Windows platform.
34931 - The current version 4.2 Linux install package does not allow you to change to the installation location to a custom location.
None.
34976 - The connectors Pervasive PSQL v10 and Pervasive PSQL v9 are not supported on Linux.
Run any profile that uses a Pervasive PSQL v10 or v9 connection on a supported Windows platform.

Important Notes

Oracle8 Connectors

Two versions of the Oracle8 connector are included with Data Profiler to support two versions of Oracle8. The djoci8.8.0.5.so connector works with all Oracle 8.0.x client installations. The djoci8.8.1.5.so connector works with all Oracle 8.1.x client installations. You must either copy or make a symbolic link from the vendor shared library to the djoci8.so shared library in the Data Profiler installation directory.

By default, the 8.1.x version of the Oracle8 connector is installed.

To enable Oracle 8.0.x support, type the following command in the Data Profiler installation directory:

cp djoci8.8.0.5.so djoci8.so

To enable Oracle 8.1.x support, type the following command in the Data Profiler installation directory:

cp djoci8.8.1.5.so djoci8.so

ODBC Drivers

Support for both the Data Direct Technologies ODBC driver manager and the IODBC driver manager are included in the installation.

By default, the ODBC 3.x connector named djodbc3.so is built for the Data Direct Technologies ODBC driver manager. The IODBC version of the ODBC 3.x connector can be used in situations where the default ODBC 3.x connector does not work with ODBC drivers from other vendors. The IODBC enabled shared library is named djodbc3.iodbc.so; the Data Direct Technologies shared library is named djodbc3.datad.so.

Useful Links

Technical Support

If you encounter problems not covered in this document, see the file contact.htm with the product documentation under /opt/PervasiveSoftware/DataProfiler<version>/help/pdf.

Disclaimer

PERVASIVE SOFTWARE INC. LICENSES THE SOFTWARE AND DOCUMENTATION PRODUCT TO YOU OR YOUR COMPANY SOLELY ON AN "AS IS" BASIS AND SOLELY IN ACCORDANCE WITH THE TERMS AND CONDITIONS OF THE ACCOMPANYING LICENSE AGREEMENT.

PERVASIVE SOFTWARE INC. MAKES NO OTHER WARRANTIES WHATSOEVER, EITHER EXPRESS OR IMPLIED, REGARDING THE SOFTWARE OR THE CONTENT OF THE DOCUMENTATION; PERVASIVE SOFTWARE INC. HEREBY EXPRESSLY STATES AND YOU OR YOUR COMPANY ACKNOWLEDGES THAT PERVASIVE SOFTWARE INC. DOES NOT MAKE ANY WARRANTIES, INCLUDING, FOR EXAMPLE, WITH RESPECT TO MERCHANTABILITY, TITLE, OR FITNESS FOR ANY PARTICULAR PURPOSE OR ARISING FROM COURSE OF DEALING OR USAGE OF TRADE, AMONG OTHERS.

© Copyright 2008 Pervasive Software Inc.

All Rights Reserved.

Pervasive Software Inc.
http://www.pervasive.com
12365 Riata Trace Pkwy, Bldg B
Austin, TX 78727 USA
Voice: (512) 231-6000
Fax: (512) 231-6010
Online Pervasive Contacts

*** END OF README ***