Home
News
Meta Integration 06/2008 - Meta Integration releases
new Web 2.0 enabled MIMB 6.0
Site Map
Home < width=46 height=34>

Meta Integration® Repository (MIR) Server
"Metadata Management" Solution

< Description | Readme | Architecture width=17 height=22>

README

Table of Contents

  1. Copyright Notice
  2. Release Changes
  3. MIR Server Installation
    1. Installing on Windows 2000, XP, or 2003 Server
    2. Installing on Unix / Linux x86
    3. Java security aspects
    4. License key
      1. On Windows 2000, XP, or 2003 Server
      2. On Unix / Linux x86
    5. Uninstalling/Re-Installing
  4. Database Setup
    1. Setting up Microsoft Access on Windows
    2. Setting up Microsoft SQL Server on Windows
    3. Setting up NCR Teradata on Windows
    4. Setting up Oracle on Windows
    5. Setting up Oracle on Unix / Linux x86
    6. Setting up IBM DB2 UDB on Windows
  5. MIR Server Upgrade
  6. Administration
    1. Starting the Server
    2. Login onto the Server
    3. Populating the database with the graphical interface
    4. Populating the database with the command line
    5. Modifying the MIRSetup.xml file
    6. Using the RepositoryBridge utility
    7. Repository Performance Tuning
    8. Automation Using Batch Scripts
  • Frequently Asked Questions
  • Technical Support

    • 1. Copyright Notice

    Copyright © Meta Integration Technology, Inc. 1997-2006.
    All Rights Reserved.
    Meta Integration® Model Bridge (MIMB), Works (MIW), Repository (MIR), Repository SDK (MIRSDK), and Database SDK (MIDBSDK) are trademarks of Meta Integration Technology, Inc.
    Other product and company names (or logos) mentioned herein may be the trademarks of their respective owners.
    http://www.metaintegration.com

    2. Release Changes

    For a full list of changes on this release, you should also have a look at:
       the MIW Readme file for the repository client application's improvements
       the MIMB Readme file for model bridges improvements
       the MIRWeb Readme file for repository web frontend's improvements

    3. MIR Server Installation

    The Meta Integration Repository and Works software can be installed:

    This document explains how to install and configure both the database server and the repository application server.

    3.1. Installing MIR on Windows 2000, XP, or 2003 Server

    Pre-Installation Check-list

    Installation

    The installation of the server for the MIR Works Personal Edition and for the MIR Enterprise Edition is the same except otherwise stated below.

    Note: MIR no longer requires a separate installation of Sun Java™ 2 Runtime Environment (J2RE)

    Setup stores various information in the file MIRSetup.xml (see Modifying the MIRSetup.xml file):

    On Windows NT and for the Enterprise Edition only, Setup installs MIR Server as a service of the System Account. the service starts automatically when the machine is rebooted. Setup may modify the system environment variable PATH to add a path to the directory containing the Java JNI library. In this case, you must configure and populate the database, then reboot the machine to start the service. Note that the server will fail to start if the database server is not started when the MIR service starts.

    Database Setup :

    3.2. Installing MIR on Unix / Linux x86

    First, install Sun Java™ 2 Runtime Environment (J2RE) version 1.5 for Unix / Linux x86.
    Note that the selected JVM must support native threads (1.5 supports them).

    Then install the Meta Integration Repository application server:

    You must now configure and populate the database before starting MIR Server, see Setting up Oracle on Unix / Linux x86.

    3.3. Java security aspects

    This product is based on the Java 2 Platform. This Platform uses a much more sophisticated security system so that the MIR Server will need a policy file in order to function correctly. This file is supplied with the MIR Server, please do not delete or modify it. The directives in this file give our Java classes access to the network and the local files. This file specifically gives these privileges to our classes and to the JavaHelp classes only.

    3.4. License Key

    The server requires a license key in order to run. This section explains how to get a license key and how to install it.

    3.4.1. On Windows 2000, XP, or 2003 Server

    The "License Key Install" program retrieves some information that we need to provide you a license key. It also allows you to install or udpate the license key once you have received it.

    If you don't have a license key yet:

    We will send you back a license key via email.

    You can install a license key by following these steps:

    Restart the server for it to take into account the new key.

    3.4.2. On Unix / Linux x86

    The License Key is stored in a file MIRServer.lic located in installation directory/bin.

    If you don't have a license key yet:

    We will send you back a license key via email.

    Copy this license key in the license key file installation directory/bin/MIRServer.lic, in place of the "Trial Version" text. Restart the server for it to take into account the new key.

    3.5 Uninstalling/Re-Installating

    Should you run into errors on install on a Windows environment, especially when a previous version is already installed, please check to be sure you are complying with the following:

    Important: When you uninstall the previous version of the MIR Works Personal Edition, please make sure you don't delete your database; answer No to the question:
      "A file MIR.mdb may already exist in . Do you want to install the database files anyway?"

    When re-installing, you will need to install MIRW in the same directory as the previous installation.

    4. Database Setup

    4.1. Setting up Microsoft Access on Windows

    The access database file mir.mdb is located in the installation directory\Repository\Database directory. This database file contains the database schema as well as some sample models and tutorial mappings. There is no further installation necessary as far as the Microsoft Access database is concerned.

    If you so desire, you may copy the file to another directory, since the size of this file depends upon the number of models that are to be placed in the database. In this case, you will need to Modify the ODBC Data Source Name configuration.

    Configuration of the ODBC DSN for MS Access

    The MIR Server installation should have created an ODBC User Data Source Name for the MIR Server to access the Microsoft ACCESS database. If this operation failed or if you want to create a new Data Source Name you can configure ODBC by following these guidelines. You will also need to modify the MIRSetup.xml file (see Modifying the MIRSetup.xml file). Note that your ODBC Administration tool might have a slightly different interface than the one described here (depending on the version of the OS and the version of the Access ODBC driver).

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or the ODBC data source administrator.
    3. Select the User DSN tab in order to create the new Data Source Name.
    4. Select the Add option to create the new DSN.
    5. Select the Microsoft Access Driver from the list of drivers and click on Finish.
    6. Fill in the DSN name that was supplied at installation in the Data Source Name field (or the name of the new DSN you want to create).
    7. In order to select the database click on the Select button, go to the location of the mir.mdb file, select it and then click on OK.
    8. If a password and username were supplied at installation you should click on Advanced and fill in the two fields, then click on OK.
    9. The configuration of the ODBC driver for Microsoft Access is now complete, click on OK to validate the changes.

    If you want only to modify the location of the database file:

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or the ODBC data source administrator.
    3. Select your Data Source Name in the User DSN tab and click on the Configure button.
    4. Click on the Select button, go to the location of the mir.mdb file, select it and then click on OK.
    5. Click on OK to validate the changes.

    4.2. Setting up Microsoft SQL Server on Windows

    This section describes how to set up SQL Server 6.5. If you are using SQL Server 7.0 or SQL Server 2000 ask your Database administrator or contact us (see Technical Support below).

    System Requirements

    MS SQL Server database creation

    In order to modify the following parameters you will need to start the SQL Enterprise Manager and log in as a database administrator. Then follow the guidelines below to configure your database:

    1. First, increase the amount of memory to 16384. This is an approximate amount which depends on the number of applications using the server concurrently and the average size of a model in the database. To increase the amount of memory select the server that will contain the MIR repository and then select the Configure SQL Server icon which is the third icon in the upper left hand corner. Then select the Configuration tab and look for the memory field in the Configuration column. Type 16384 into the Current column.
    2. You must also increase the number of locks that are available to 10000. In the same window look for the field locks in the Configuration column and then change the value in the Current column to 10000. Be very careful that the correct amount of memory is allocated when increasing the locks or you may have problems restarting the server.
    3. After having made these modifications select Apply Now followed by OK. Now stop and restart the server by first selecting the stop light icon. To stop the server select the red stop light and wait for the server to stop. To restart the server, select the green light.
    4. Now create a database device for the model metadata. To do this select the Database devices menu item in the Manage menu. At this point a display will show the amount of free disk space. The New device icon should then be selected. Then supply the name of the new data device and its size. In general, a size of 500 megabytes should be sufficient for about 100 models with around 50 classes each.
    5. Next create the database by selecting the Database menu item in the Manage menu. At this point the amount of free space in each database device is displayed. Select the New database icon and supply a name for the database, choose the database device created in step 4 to contain the database, and set the size of the database equal to the size of the database device. The selection of a log device is not necessary.
    6. The amount of tempdb should also be increased if you will be storing models with more than 40 classes. To do this expand the server tree and then expands the database tree list. Right click on the tempdb database and select the Edit item. A new window will appear and you should then select the Expand button. Select a data device that has room for an expansion of 50 megabytes and then click on the Expand Now button. If you need another database device to allocate this additional space, create a new device by selecting new device and fill out the window as described in step 4.
    7. Now the database is ready to receive the schema and the stored procedures that are necessary for the correct functioning of the MIR database. The schema for the database is located in the Repository\Database directory of the installed product. To install the schema expand the server tree and then the database tree. Then select the newly created database, and select the SQL Query Tool menu item in the Tools menu. The name of the database should be in the leftmost pull down menu. Then select the load SQL script icon and a file selection box will be displayed. Navigate to the database directory in the installation directory and select the file MIR-MS-SQLServerCreate.sql. Once this file is loaded into the window click on the green arrow button to execute the DDL.
    8. If the login provided during the ODBC configuration of the MIR Server does not exist yet in the database you must now create it. To do this expand the server tree and right click on Logins to select the New Login menu entry. Then type in the DSN login name, the corresponding password, click on the Add button and finally click on the Close button.
    9. Then go back to the database you have just created and expand it to select Groups/Users. Right click and select the New User menu entry. Type in the user name and the login you have just created. Then click on the Add button and then the Close button.
    10. The user you have just added to this database should now be visible in the tree.
      1. For SQLServer versions earlier than 2003 right click on it and select the Permissons menu entry. Then Click on the Grant All button, on the Set button and finally on the Close button.
      2. For newer versions of SQL Server expand the server tree and the Logins tree, right-click the user you created earlier and select Properties, select the Database Access tab, and select the database you created earlier, and under database Roles assign the public, db_datawriter and db_datareader roles.

    The database is now configured.

    A file is also supplied for removing the schema, MIR-MS-SQLServerDelete.sql.

    If you are upgrading from version 2.5.x or 3.0.x to version 3.1, see MIR Server Upgrade to know how to migrate your data.

    Configuration of the DSN for MS SQL Server

    The MIR Server installation should have created an ODBC System Data Source Name for the MIR Server to access the Microsoft MS SQL Server. If this operation failed or if you want to create a new Data Source Name you can configure ODBC by following these guidelines. You will also need to modify the MIRSetup.xml file (see Modifying the MIRSetup.xml file).

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or the ODBC data source administrator.
    3. Select the System DSN tab (or the User DSN tab if you cannot create System DSNs) in order to create the new Data Source Name. Note that, if you selected the User DSN tab, you won't be able to start the server as a service.
    4. Select the Add option to create the new DSN.
    5. Select the SQL Server driver from the list of drivers and click on Finish.
    6. Fill the fields of the creation window. Depending on your installation, the ODBC Administrator may instead display a Wizard that will guide you in the creation of new Data Source Name.
      1. Fill in the DSN name that you supplied at installation time in the Data Source Name field (or the name of the new DSN you want to create).
      2. Type in the description if desired.
      3. Then select the server that contains the database. If the server is not in the list then type in the name of the server.
      4. Next click on the Options button in order to enter the name of the database that was created for MIR Server.
    7. Validate everything by clicking all the OK buttons.

    After having completed the fields, select the OK button.

    If an error such as "unable to find the table mirproduct" occurs while starting the MIR Server, it is probably due to the fact that the database field above was incorrectly filled.

    4.3. Setting up NCR Teradata on Windows

    This section describes how to set up NCR Teradata on Windows. This is only an overview of the tasks to perform.You should also contact your Database Administrator for the tuning of the database.

    System Requirements

    NCR Teradata database creation

    In order to modify the following parameters you will need to start the WinDDI and log in as a database administrator. Then follow the guidelines below to configure your database:

    1. Start the WinDDI and log in as a database administrator.
    2. Make sure the user you provided during the installation exists or create it now.
    3. Create a Database.
    4. Make sure the user provided during the installation has read-write-create access to the database
    5. Quit the tool WinDDI.
    6. Start the QueryMan tool and log in using the DSN generated during the installation.
    7. Load the SQL query file MIR-NCR-TeradataCreate.sql located in the installation directory\Repository\Database directory.
    8. Select Execute in the Tools main menu to execute the queries one statement at time.
    9. Quit QueryMan.

    The database is now configured.

    If you are upgrading from version 2.5.x or 3.0.x to version 3.1, see MIR Server Upgrade to know how to migrate your data.

    Configuration of the DSN for NCR Teradata

    The MIR Server installation should have created an ODBC System Data Source Name for the MIR Server to access the NCR Teradata Database. If this operation failed or if you want to create a new Data Source Name you can configure ODBC by following these guidelines. You will also need to modify the MIRSetup.xml file (see Modifying the MIRSetup.xml file).

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or the ODBC data source administrator.
    3. Select the System DSN tab (or the User DSN tab if you cannot create System DSNs) in order to create the new Data Source Name. Note that, if you selected the User DSN tab, you won't be able to start the server as a service.
    4. Select the Add option to create the new DSN.
    5. Select the Teradata driver from the list of drivers and click on Finish.
    6. Fill the fields of the creation window.
      1. Type in the DSN name that you supplied at installation time in the Data Source: Name field (or the name of the new DSN you want to create).
      2. Type in the description if desired.
      3. Type in the server name that you supplied at installation time in the Teradata Server Info: Name field.
      4. Type in the name of the Repository Database in the Default Database field.
      5. Type in the name of the user used to connect to the database in the Username field.
      6. Optionally, type in the password of the user used to connect to the database in the Password field. Note that the MIR Server will use the user information provided in MIRSetup.xml file to connect to the database (not the information provided in the DSN).
      7. Next click on the Options button:
        • Change the Session Mode to be "ANSI".
        • Change the DataTime Format to be "AAA".
        • Check with your Database Administrator if other fields should be modified.
    7. Validate everything by clicking all the OK buttons.

    After having completed the fields, select the OK button.

    If an error such as "unable to find the table mirproduct" occurs while starting the MIR Server, it is probably due to the fact that the Default Database field above was incorrectly filled.

    4.4. Setting up Oracle on Windows

    System Requirements

    Oracle database creation

    First install the Oracle Database software on your computer. Please refer to the Oracle installation documentation if needed.
    This installation process should install on your computer:

    When the Oracle Database server software is installed, you can either

    You may need to manually start the following Windows NT services:

    Once the Database instance is properly running on the Oracle server, you should create a new Oracle USER <the user> who will be the owner of the MIR schema.
    Please use the DBA Studio database administration tool in order to do so.
    Make sure this user has the following Roles: CONNECT, DBA, RESOURCE.

    Then you should configure your TNS Net Service Name (if necessary) and create an ODBC DSN pointing to this database. The ODBC DSN should use <the user> name and password (see next paragraph).

    You should now test that both the ODBC DSN and the TNS Net Service Name are properly set up by trying to connect to the database.
    To try the ODBC connection, you should use the Oracle-ODBC-Test software and <the user> name and password.
    To try the TNS Net Service Name you should try to connect to the database using the SQL-Plus software and <the user> name and password.

    Now, you can install the MIR schema:

    The database is now configured.

    A file is also supplied for removing the schema, MIR-MS-OracleDelete.sql.

    If the MIR Server installation did not create a DSN, create it now.

    If you are upgrading from version 2.5.x or 3.0.x to version 3.1, see MIR Server Upgrade to know how to migrate your data.

    Configuration of the ODBC User Data Source Name

    The installation procedure is divided into two parts. First you need to create a TNS Net Service Name.

    1. Start the Net8 Configuration Assistant
    2. Choose Local Net Service Name Configuration and click on Next.
    3. Choose Add and click on Next.
    4. Choose the appropriate version of Oracle database you want to access and click on Next.
    5. Type in the service name or SID of your database and click on Next.
    6. Choose the network protocol and click on Next.
    7. Type in the hostname of the machine on which the Oracle database is installed and choose a port, click on Next.
    8. Perform a test if you want to validate the information you entered and click on Next.
    9. Enter the Net Service Name for this installation and click on Next.
    10. Choose "No", click on Next two times then click on Finish.

    Now you need to create a new Data Source Name.

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or ODBC data source administrator.
    3. Select the System DSN tab (or the User DSN tab if you cannot create System DSNs) in order to create the new Data Source Name. Note that, if you selected the User DSN tab, you won't be able to start the server as a service.
    4. Select the Add option to create the new DSN.
    5. Select the Oracle driver from the list of drivers and click on Finish.
    6. Fill the fields of the creation window.
      1. Type in the DSN name that you supplied at installation time in the Data Source: Name field (or the name of the new DSN you want to create).
      2. Type in the description if desired.
      3. Type in the Net Service Name that you supplied in TNS Service Name creation.
      4. Depending on the ODBC driver, you may have to enable the option Enable Closing Cursors
    7. Validate everything by clicking on the OK button.

    4.5. Setting up Oracle on Unix / Linux x86

    System Requirements

    Oracle database creation

    The following guidelines are for Oracle 7.3.
    If you want to install on a newer version of Oracle, we recommend that you ask your DBA to do it for you.

    The first thing to do is to create a table space for the MIR metamodel schema. The size of the table space depends on the average size and number of models that are going to be placed into the repository. It is difficult to predict the size of the required database, but on average 500 megabytes should be sufficient for about 100 models with around 50 classes each.
    Here is the list of modifications that need to be performed.

    1. The parameter shared_pool_size should be set to 28000000. To modify this value you should edit the file DatabaseName.ora (where DatabaseName is the Oracle database name that is provided at installation time or by the administrator) found in the dbs directory located in the Oracle Home directory. One should be sure to increase the shared memory parameter in the /etc/system file so that there is a sufficient amount of shared memory available. This is done by adding/editing the line set shmsys: shminfo_shmax, this value should have already been set at the installation of Oracle.
    2. Create the table space for the model metadata information while logged in with the "dba" privilege, the command would be the following for a 500 megabytes database. You may wish to consult the Oracle manual for this command since it may change between releases.
      CREATE TABLESPACE MIR DATAFILE 'mir.ora' SIZE 500M DEFAULT STORAGE ( INITIAL 300M NEXT 200K MINEXTENTS 3 PCTINCREASE 5 );
    3. Depending on the size of the largest models in the database you might need to increase the size of the rollback segments. A size of 1 megabyte is recommended. If a table space is not available for the rollback segment you should create a tablespace as described in step 2.
      CREATE PUBLIC ROLLBACK SEGMENT R05 TABLESPACE "LeftOver" STORAGE ( INITIAL 3M NEXT 128K MINEXTENTS 3 );
    4. You should then log in as the database administrator and create a user that will use this table space. This user should have DBA privileges in this table space. The user name must be the same as the one provided during the installation of MIR Server. You should also create a temporary table space for the MIR table space if you are going to place large models into the repository.
      CREATE USER jack IDENTIFIED BY lollypop DEFAULT TABLESPACE MIR TEMPORARY TABLESPACE TEMPORARY_DATA PROFILE DEFAULT;
      GRANT CONNECT TO jack;
      GRANT DBA TO jack;

    Configuration of the ODBC User Data Source Name

    The ODBC connection to the database is configured automatically at installation time with the information provided in the installation script. If you would like to change a configuration parameter, the files are located under the conf directory of your Meta Integration installation directory.

    The Database Source Name (DSN) is configured in the odbc.ini file: 

    [MITI-MIR]
Driver=/opt/oracle/instantclient/libsqora.so
ServerName=miroracle
UserID=
UserPassword=

    The server name is configured in the tnsnames.ora file: 

    miroracle=(DESCRIPTION=
	(ADDRESS=
		(PROTOCOL=TCP)
		(HOST=_machine_)
		(PORT=1521))
	(CONNECT_DATA=
		(SID=_sid_)))

    Testing the ODBC connection

    To test the ODBC connection to the Oracle Database, you can use the scripting command line utility: 

    ./MIR.sh -l

    4.6. Setting up IBM DB2 UDB on Windows

    System Requirements

    DB2 UDB database creation

    First install the DB2 UDB Database software on the database server computer (first tier). Please refer to the DB2 UDB installation documentation if needed. This installation process should install:

    Then, create a new database (DBALIAS) called MIR in the current DB2 instance to isolate the MIR database from other databases that may be already running on this database instance. You can use the DB2 Control Center wizard to create this new database.

    Then, create a DB2 schema called MIR which will hold the MIR tables and views. Create a DB2 user with sufficient access rights to access this schema. You can use the DB2 Control Center tool to do this.

    Then, open the MIR SQL DDL script in a text editor to review it.
    By default, this file is located in C:\Program Files\MetaIntegration\Repository\Database\MIR-IBM-DB2Create.sql
    This script will create a MIR BufferPool and Regular MIR TableSpace which size is larger than the default DB2 settings.
    This configuration is required, because some of the MIR Tables and Views have many large columns.
    Please review the MIR BufferPool and Regular MIR TableSpace creation script, and feel free to customize it to your needs.

    Alternatively, the MIR BufferPool and MIR TableSpace can be created manually using the DB2 Control Center wizard as follows:
    (If you do so, please comment or remove the first three lines of the DDL script of course)

    Now, you can run the SQL DDL script to create the MIR Tables and Views:

    It is necessary to setup the transaction log to allow for import of models into the system.

    The database server is now configured.

    A file is also supplied for removing the schema, MIR-IBM-DB2Delete.sql.

    During the MIR application server installation process (second tier), an ODBC DSN pointing to this database will be automatically created using the DBALIAS and user (name and password) information supplied in the installation wizard.
    If you selected "IBM DB2" during installation, an ODBC DSN should have already been created.
    You can also configure the ODBC DSN manually, as described below.

    Configuration of the ODBC User Data Source Name

    To create a new Data Source Name:

    1. Activate the Control Panel window.
    2. Open the 32bit ODBC or ODBC data source administrator.
    3. Select the System DSN tab (or the User DSN tab if you cannot create System DSNs) in order to create the new Data Source Name. Note that, if you selected the User DSN tab, you won't be able to start the server as a service.
    4. Select the Add option to create the new DSN.
    5. Select the IBM DB2 ODBC DRIVER driver from the list of drivers and click on Finish.
    6. Fill the fields of the creation window.
      1. Type in the DSN name that you supplied at installation time in the Data source name field (or the name of the new DSN you want to create).
      2. Type in the description if desired.
      3. Select the Database you want to connect to in the Database Alias field.
    7. Click on the OK button.
    8. Edit the definition of the newly created DSN by clicking on the Configure... button.
    9. Fill the User ID field with the user name information
    10. Fill the Password field with the user password information
    11. Click on the Connect button to check the connection is working.
    12. Click on the Advanced Settings tab:
      1. Make sure the value of CLI parameter DBALIAS is set to the database alias you want to connect to.
      2. You can also set the value of CLI parameter CURRENTSCHEMA to the schema name to be used by, if it is different from the default schema for this user.

    5. MIR Server upgrade

    The MIR Works Personal Edition Version 3.1 is delivered with a Upgrade utility that let you migrate your Database (version 2.5.x or 3.0.x) to Version 3.1.0:

    The MIR Enterprise Edition Version 3.1 is delivered with a Upgrade utility that let you migrate your Database Version 3.0.x to Version 3.1.0:

    Please contact us at for the migration of your Enterprise Edition repository database (Oracle, SQL Server or Teradata) if you are upgrading from release 2.5.

    Migration procedures for migrating from the release 2.4.x or older are freely available upon request. Please use the installation procedures in the "MIR Server Installation" section then contact us at for the migration of the old database.

    6. Administration

    6.1. Starting the Server

    On Windows

    In the MIR Enterprise Edition, the MIR Server is installed so that it is started automatically. On Windows NT it is installed as a service. On Windows 9x it is installed as a shortcut in the Startup folder.

    The MIR Server program may be started as an application (if is not installed so as to start automatically or if the program has been stopped). You can start it by opening the "Start Menu", going to the "Meta Integration" folder under "Programs", and clicking on the "Repository Server" entry. You can also start it from the command line by typing "MIRServer -run" (supposing that you have set the PATH accordingly).

    The client product is also installed by this Setup. To know how to start the client, see the works client Readme file.

    The MIR Server program may also be started as a stand-alone version (Personal Edition). Setup creates a command batch file named MIRW.bat in the installation directory (in the 'bin' subdirectory). It also creates a shortcut to this file in the "Start Menu".
    You start the application by opening the "Start Menu", going to the "Meta Integration" folder under "Programs", and clicking on the "Repository Works" entry.

    On Unix / Linux x86

    If the user installing the product is the root user, the MIR Server is installed so that it is started automatically using the System V init scripts.

    The MIR Server program may be started as an application (if is not installed so as to start automatically or if the program has been stopped). The command is RepositoryServer (supposing the PATH variable has been set to contain installation directory/bin).

    6.2. Login onto the Server

    The server requires a user name and a password. The installation default administrator's login user name is 'administrator' and password 'administrator'. The installation default user name is 'anonymous' and the default host is 'localhost'.

    6.3. Populating the database with the graphical interface

    The models are stored in a hierarchy composed of Products, Versions and Formats. Thus populating the database consists in creating these elements in turn using the graphical interface.

    When you right click on a Product, Version or Format object the popup menu will contain an Add option allowing you to create new objects in the database.

    When one of the add option is selected you will be prompted for the name of the new object. Once the object is created you can modify its properties in the attribute panel.

    From there, there are two ways to attach a model to a "Model Metadata" Format object.

    The first solution is to use Meta Integration Model Bridge to convert a model to the MIM format. This file can then be opened in Meta Integration Works, and "copy and pasted" to the Format object.

    The second solution is to use the server to convert the model. Right click on the Format object, and select the Import entry. In the displayed dialog box, you can choose your modeling tool among a list of available tools. You can then select the file containing the model. It will then be sent to the server, converted to the MIM format and put in the database.

    6.4. Populating the Database with the command line

    The models are stored in a hierarchy composed of Products, Versions and Formats. Thus populating the database consists in creating these elements in turn using the tools listed below. See the Meta Integration Repository meta model for more details.
    You can also use Meta Integration Works to create, delete and modify these elements (you must log in as the "administrator" user).

    In the command descriptions below the parameters that are in bold are mandatory.

    MIR - This command creates a product,version or format in the database.

    MIR -a repository_object_type -p repository_path

    -a | --add repository_object_type
    the type of object to add to the repository: Product, Version, ModelFormat
    .
    -p | --path repository_path
    the path of the object in the repository. This path is absolute: it always starts from the root of the repository and finishes by the name of the object to manipulate.

    An example file as well as example metadata for Rational's Rose, CA's ERwin, etc. is supplied with the MIR product. This file is called MIRCreateSampleModels.bat on the Windows platforms and MIRCreateSampleModels.sh on the Unix / Linux x86 platforms.

    After having populated the database start MIRServer. To view the information, use the MIW client.

    6.5. Modifying the MIRSetup.xml file

    The MIRSetup.xml file contains information that is necessary for the correct functioning of the MIR server application, as well as the MIW Client. In particular, it provides the parameters of the database connection. memory size specifications, and the location of the class files to be used by the Java run time environment. The installation procedure initializes this file so that you should not have to modify it unless you want to change the location of the log file or use the server with specific parameters.

    The file is self-documenting. Each parameter is given an XML tag, definition, usage and examples within the text of the file.

    This file resides in the MetaIntegration\conf directory.

    The MIRSetup.xml file is located in the conf directory.

    This file can be modified in any text editor, and is autodocumented using xml comments.

    The MIRSetup.xml file contains three sections underneath Setup:

    Directory
    for designating the installation directories.
    Home
    for designating the main Meta Integration directory.
    Bin
    for designating the binary execution directory.
    Log
    for designating the log directory.
    Temp
    for designating the temp directory.
    Java
    Meta Integration Java Environment Setup.
    Directory
    Meta Integration JAR file Directory.
    ClassPath
    Defines the Java Class Path.
    AppendSystemClassPath
    Specifies wether to append the system ClassPath.
    LibraryPath
    Defines the Java Library Path.
    VirtualMachinePath
    Defines the Java Virtual Machine (JVM) Path.
    RunTime
    Java Run Time Environment (JRE or JDK).
    Executable
    Purpose: JRE or JDK execution Options.
    Server
    Meta Integration Server Configuration.
    Parameters
    MIR Server Execution Paramaters.
    Encryption
    MIR Client/Server Communication Encryption.
    Client
    Meta Integration Client Setup.
    Directory
    MIR Client Execution Paramaters.
    Repository
    Meta Integration Repository Configuration.
    Directory
    Identify the MIR base directory.
    Database
    Defines the MIR Database Connection.
    DataSourceName
    Identify the MIR Database Name.
    LoginUser
    Identify the MIR Database Login User.
    LoginPassword
    Identify the MIR Database Login Password.
    ModelBridge
    Meta Integration Model Bridge Information.
    XmlConversationDirectory
    Identify the XML Conversation Directory.
    License
    Meta Integration Licensing Information.
    Mode
    License Key Mode.
    Key
    License Key specific to your hostid.
    Server
    Defines the MIR License Server connection parameters.
    Name
    MIR License Server name or address.
    Port
    MIR Licenser server port number.

    Note that these keywords are case sensitive.

    Note: Some metadata bridges require shared libraries supplied by the vendors to be accessible at runtime on the computer where the bridge is executed.
    Other bridges may require the complete Design Tool to be installed.

    For example:
    The Sterling Software CoolGen Export Bridge requires the "Workstation Update API" to be installed.
    The Sterling Software CoolGen Import Bridge requires either the "Workstation Read Only API" or the "Workstation Update API" to be installed.

    6.6. Using the RepositoryBridge utility

    MIR is delivered with a RepositoryBridge utility that lets you migrate your repository data between an Access database and your RDBMS Database (Teradata, SQL Server, Oracle). This utility allows you to export your repositoy data to a shareable Access file or import them for an MS Access file.

    Import from MS Access into your repository:

    Warning: the current data in your repository will be deleted and replaced by the one from the MS Access repository.

    To import your repository data from the MS Access (.mdb) file, just run the "RepositoryBridge import" utility in the Start menu "Programs\MetaIntegration".
    This program will open a DOS window and guide you through the import process.
    Once done, the details of the import process can be seen in the RepositoryBridge log file (log/RepositoryBridgeLog.txt).

    Export from your repository to an MS Access file:

    To export your repository data into an MS Access (.mdb) file, just run the "RepositoryBridge export" utility in the Start menu "Programs\MetaIntegration".
    This program will open a DOS window and guide you through the export process.
    Once done, the details of the export process can be seen in the RepositoryBridge log file (log/RepositoryBridgeLog.txt).

    6.7. Repository Performance Tuning

    MIR is delivered and initially configured for minimal footprint in installation and memory requirements. In order to maximize performance and availablity of the server there are several ways to adjust the systems. In particular, it is important to ensure proper database configuration, proper repository server memory allocation and utilization, and proper connectivity between the repository server and database server (if they are separate).

    Database Performance Tuning

    In order to understand how to properly configure the database for efficient access by the repository, it is important to understand that the MIR does NOT behave like most applications that databases are tuned for. Specifically, MIR is NOT A TRANSACTION HEAVY application. It is QUITE DIFFERENT from an OLTP, Data Warehouse, or other transaction based system. Except for the web metadata search query, it is QUITE DIFFERENT from OLAP applications as well. This fact must be clearly communicated to the Database Administration staff.

    Instead, MIR mostly reads and writes ENTIRE MODELS or CONFIGURATIONS (collections of model and mapping formats) at a time, then does nothing in terms of database access. Thus, in order to maximize performance, it is necessary to maximize buffer, cashing, cursor, etc., sizes, since the are used intensively in a short period of time. For specific tuning suggestions, please refer to the documentation of the specific database being deployed, and tune the parameters for bulk SQL reads and bulk SQL updates.

    Repository Server Performance Tuning

    Given a fixed clock speed for the repository server, the greatest impact on performance is based upon memory availablilty and utilization. The MIR uses a single configuration file for all startup options, it is at Meta Integration\conf\MIRSetup.xml. The setting is defined by the <Options> XML tag contained in the <Setup> and <RunTime> containers in the XML file. As MIR is a Java application, the memory settings are defined by the Java startup parameters -Xms(size) and -Xmx(size), defining the initial and maximum memory allocation, e.g.:

    <Options>-Xms256m -Xmx512m</Options>

    If you are also running the MIRWeb server on this same machine, be sure not to make the memory requirements so large that MIR is competing with the web server for resources.

    Connectivity Between Servers

    While it should go without saying, the connectivity, expecially of large, bulk, quantities of data (an entire model at a time), between the database server and the repository server machine is a very critical factor in performance. Your systems administration staff should be able to diagnose such problems.

    Automation Using Batch Scripts

    All non-interactive functionality provided by Meta Integration Works (MIW) can be reproduced using command line calls to MIR.exe and MIMB.exe (on DOS machines). On UNIX the commands are MIR and MIMB.

    MIR

    MIR allows one to perform all activities, including add, delete, set properties, create reference links and other repository structure based activities (those accessible from the model manager), based upon the following command signature:

    Usage:
    [-l|--list
    [-a|--add
    [-r|--remove

    [-f|--file
    [-t|--type
    [-v|--value
    [-d|--description

    [-g|--get
    [-s|--set

    [-p|--path
    [-q|--dstPath

    [-e|--export Script|Configuration|Repository

    [-u|--userName
    [-o|--oldPassword
    [-n|--newPassword

    [-b|--expiredAttributeName
    [-c|--changedSince

    [-z|--stitch [ fromSource | fromDestination ]
    [-m|--migrateFrom

    = [ |
    ]
    = [ Product | Version | ModelFormat |
    MappingFormat | FileFormat ]
    = .
    = ..
    = [ String | Number | Date ]

    Sample:
    MIR --add Product.Level --type Number --value 1 --description "Required Level to access this Product."
    MIR --set Product.Level --path P1\P2 --value 12
    MIR --get Product.Level --path P1\P2

    It is possible to add constraints to a property type:
    [min,max]: min/max value, min or max can be empty
    MIR --add Product.Min --type Number[12, ] --value 12
    MIR --add Product.Max --type Number[ ,42] --value 12
    MIR --add Product.MinMax --type Number[12,42] --value 12 ([Name=]Value,...): enumerated values
    MIR --add Product.ColorS --type String(Red ,Green ,Blue ) --value Red
    MIR --add Product.ColorN --type Number(Red=0,Green=1,Blue=2) --value 0

    Updating password:
    MIR -u userName -o oldPassword -n newPassword

    Stitching models:
    MIR --stitch --path Source\1\Model --dstPath Destination\1\Model

    Checking for triggers:
    Triggers can be checked on both --list and --export
    MIR --list --expiredAttributeName Product.ExpirationDate
    MIR --export --changedSince "20061212"

    MIMB

    MIMB allows one to execute any bridge (import or export) based upon the following command syntax:

    usage: MIMB.exe [-?|--?] [-h|--help]
    [-i|--import (import bridge name: CaErwin3Erx, ...)]
    [-f|--from (file or directory to import from)]
    [-o|--importOption (import bridge option: name=value) ...]
    [-s|--subset (P1/P2/D1, PACKAGE_FILES, DIAGRAM_FILES)]
    [-e|--export (export bridge name: IbmRationalRose98, ...)]
    [-t|--to (file or directory to export to)]
    [-p|--exportOption (export bridge option: name=value) ...]
    [-l|--logFile (MIMBAction::log file: stderr by default)]
    [-c|--consistency (validation level : 0, 1 ,2)]
    [-v|--validation (validation level: NONE, BASIC, ADVANCED)]
    [-r|--request (request XML File)]
    [-y|--reply ( reply XML File)]
    [-m|--mapping (file to save the mapping to: xml or csv)]
    [-j|--join (INNER,OUTER_EXPORT,OUTER_IMPORT,FULL_OUTER)]

    There are numerous examples in /scripts. Additionally, use the "Script As" button in MIMB Desktop to create a batch script version of any import and export sequence that can be created using the MIMB UI.

    7. Frequently Asked Questions

    You can also look at the Frequently Asked Questions in the MIW Readme file.

    1. On Windows, when launching the self extracting executable, an "Overwrite Protection" window appears: "The following file is already installed on your system".

    This is because you have launched this executable shortly after a previous installation and the previous InstallShield Setup has not yet finish its cleanup. Click on "Cancel" and restart the program. If you click on any other button, you may experience problem during the installtion because some files would not have been copied correctly. Typically, an error (-113) will occur during the move data process. If you experience such an error, wait a few seconds then restart the installation.

    2. On Windows, why must I be an Administrator to install MIRServer?

    For the MIR Enterprise Edition for Microsoft SQL Server, Setup installs MIRServer as a service of the System Account. You don't need to be an administrator to install MIR Personal Edition.

    3. The server refuses to start. where can I get information?

    Examine the log file to determine the type of error that has occurred:

    4. Problems with the file MIRSetup.xml.

    The MIRSetup.xml file contains information about connecting to the database - DSN, user name, and password. When the system cannot connect to the database or cannot read the file it gives the same error. If the file exists you should check and make sure that the information for connecting to the database is correct.

    5. I have problem writing to the database. The error message is "unable to update a counter". Why?

    If the database is Mircrosoft ACCESS this error may result if the mdb file is read only. If the database is Microsoft SQL Server or Oracle 7.3 or 8.0, this error may appear if you don't have write access to the tables. In this case, the database administrator must give the access to the tables for the user specified in the MIRSetup.xml file. In general, this user should create the tables to avoid any problems in Microsoft SQL Server and Oracle.

    6. In the personal edition, where do I set the Classpath variable for the JDBC bridge so that it may locate my driver class?

    The directory or jar file pathname must be set on the command line in the batch file that starts the MIRServer application.

    You need to edit the batch file MIRW.bat to append the directory or jar file pathname to the '-classpath' option, and then restart the application.

    7. On windows, I cannot start the Repository Server, it complains that it cannot find jvm.dll. Why?

    On Windows NT, the error message box says "The dynamic link library jvm.dll could not be found in the specified path...".
    This may happen if you installed the product as one user and you run it using another identity.
    The installation updates the user environment variable "path" so that it contains the path to a jvm.dll library.

    On windows 9x, the error message is "A required DLL file, JVM.DLL, was not found". The installation modifies the environment variable "path" in the file "autoexec.bat". For this change to take effect, you must reboot your machine. Before to do so, make sure the installation updated the autoexec.bat file (the update may have failed because the file was read-only). The file should contain a line like:
    set path=c:\progra~1\jre\1.3\bin\hotspot;<other paths>
    Note that the jvm.dll path does not need to appear first. The pathname to the jvm.dll file varies with the version of the JRE being installed.

    8. On Windows, when Repository Server is started as a Service and I log off, the Service stops with an error. Why?

    You use J2SE version 1.3.0 or earlier. Install J2RE 1.5 and reinstall MIR.
    Because of a bug in JDK 1.3.0, the server will not work properly as a Service. This was fixed in JDK 1.3.1 but requires the option "-Xrs" to be set. The JDK 1.3.0 and earlier don't have this option.

    9. When saving a large Model into a MIR repository server based on IBM DB2, the following error occurs "SQL0964C The transaction log for the database is full".

    The size of the transaction rollback log of this DB2 database is not large enough, please contact your IBM DB2 database administrator so that he increases its size. These parameters can be changed in the DB2 Control Center tool as follows:

    10. My batch script (on Windows) ends up in an infinite loop due to a call to another script that is not referred to in the original script.

    On windows, the .bat extension is given precedence over the .exe extension. This means that if a script calls MIR.exe by using MIR (without an extension) it may instead be calling a faile named MIR.bat someplace in the PATH environment variable locations. Best practice on Windows installations is to always fully qualify the name of the MIR and MIMB commands by using MIR.exe and MIMB.exe.

    8. Technical Support

    The best way to get support is to send an email to with a detailed description of your problem, the MIR version that you use, your platform and your database.

    You can also get more information about our products by going to our Web Site, http://www.metaintegration.com or sending an email to .

    < Description | Readme | Architecture width=17 height=22>