MagicDraw TeamworkServer UserGuide
MagicDraw TeamworkServer UserGuide
18.2
user guide
No Magic, Inc.
2015
All material contained herein is considered proprietary information owned by No Magic, Inc. and is not to be
shared, copied, or reproduced by any means. All information copyright 2000-2015 by No Magic, Inc. All Rights
Reserved.
CONTENTS 0
Teamwork Server Concepts 5
Teamwork System Design 7
Native Repository 8
SVN Repository 8
ClearCase Repository 9
Comparison of Teamwork Server Repositories 10
Getting Started with Teamwork Server 10
Starting Teamwork Server 10
Starting the server using GUI 10
Starting the server without using GUI 12
Stopping Teamwork Server 15
Upgrading Teamwork Server 15
Automatic upgrade of Teamwork Server 15
Manual upgrade of Teamwork Server 16
Importing projects and users from earlier versions of Teamwork Server 19
Replacing the projects folder 19
Modifying the muserver.projects_directory property in muserver.properties file 20
Changing the repository location in Administrator’s Console 20
Migrating from the SVN/ClearCase repository to the Native repository type 21
Moving Teamwork Server 21
Managing Teamwork Server 22
Customizing Teamwork Server properties 23
Changing Administrator Password 24
Users Management 24
User permissions 27
Starting the Administrator’s Console 28
Administrator’s Console Dialog 28
Active Users tab 29
Projects tab 30
Log File tab 34
Properties tab 35
Secured Connection tab 35
Repository tab 38
LDAP Integration tab 44
Administrating Server via Command Line Utility 44
Giving commands on the input stream 46
Data Migration between Different Repositories 46
Changing Teamwork Server Debugging Mode 47
LDAP Support 48
Enabling LDAP Integration 48
Connection Settings 49
Authentication Settings 50
Authentication settings for the Simple User+Password authentication type 51
Authentication settings for the SASL authentication type 54
User Data Retrieval Settings 55
Connection Testing 56
Subversion and LDAP Integration Working at the Same Time 56
With MagicDraw Teamwork Server you can assign as many developers as needed to work simultaneously on
the same project using multiple workstations. The resulting server project is saved on the server for sharing
with other MagicDraw® applications. Users with administrator rights can create new users by creating a name
and assigning various permissions to work on projects. The permissions assigned will determine whether the
new user can update, commit, edit, create, and delete model elements, diagrams, and projects.
To enable Teamwork support, you should install and run MagicDraw Teamwork Server. Each MagicDraw
application is a client of Teamwork Server.
Author
A user who has committed a new project version.
Version
A unique number assigned to the committed project. Project version numbers begin at zero (for
the initial version) and increase with every new project version.
Comment
Optional description of changes in the committed version.
Tag
Information about the status of a project (approved, initially tested, etc.), or other important
information.
Administrator Login
The default Administrator's account in Teamwork Server is:
Login name: Administrator
Password: Administrator
To prevent illegal access, it is advisable to change the Administrator's password.
For more information, see Section “Managing Teamwork Server” on page 22.
Repository
A storage place for projects and their versions managed by the Teamwork Server.
Project category
A concept which enables visual grouping of projects in Teamwork Server repository.
Native user
A user whose account data is stored locally, i.e. in the native Teamwork Server repository.
External user
A user whose account data (all except the login name) is stored in an external database, e.g.,
Subversion, ClearCase, or LDAP.
Home server
A server where project has been initially created.
Domestic project
A project created on the home server.
Foreign project
A project transferred from its home server after synchronization. The foreign project cannot be
modified on the server, to which it was transferred, but it can be browsed, analyzed, selected for
report generation, and used in other projects on this server. The foreign project can have
domestic (editable) branches.
Clients communicate (using Java RMI, over TCP/IP) with Teamwork Server to retrieve projects stored on the
server, edit them, and commit them back to the server for storage.
The Teamwork Server keeps track of projects versions. Additionally, it performs several administrative functions
such as user login, authentication, and check permissions to access projects.
The Teamwork Server uses repositories for project version storage. The administrator can select any one of the
supported repository types in the Teamwork Administrator's console to configure the server (for more
information, see Section “Starting the Administrator’s Console” on page 28). Data can migrate from one to
another repository type. This functionality is also accessible from the Teamwork Administrator's console.
Regardless of the repository used, users will not feel the difference because the user workflow remains the
same.
For more information about specifying repositories, see the
Administrator’s Console dialog description, "Repository tab" on page 38.
For more information about importing or exporting project to the native
repository, see the Administrator’s Console dialog description, "Projects
tab" on page 30.
Native Repository
This is a default repository type. When the Teamwork Server is first installed and started, it is configured to use
this type of repository. The Native repository is the only type of repository available in the version prior to 12.5 of
Teamwork Server.
When the Teamwork Server is configured to use the Native repository, a directory is designated for project
storage. The Server then uses its internal proprietary code to implement a versioned repository for a collection
of projects.
SVN Repository
The Teamwork Server can be configured to use SVN repository as a back-end. In this mode, the Teamwork
Server retrieves and commits project versions into the SVN repository.
To use this repository type, the SVN client executable must be correctly installed on the computer where the
Server runs. The Teamwork Server must be able to launch the SVN executable, that is SVN executable must
be accessible on the system's PATH and have appropriate permissions to execute. Supported SVN client
versions are 1.4, 1.5, 1.6, 1.7, and 1.8.
Teamwork Server with the SVN repository supports pass-through authentication into the SVN. A pass-through
authentication is used for all access methods, except the file:// method. In this case, the Teamwork Server only
maintains a list of users but not their passwords, their passwords are not stored. When a user tries to log on the
Teamwork Server, server does not verify the password itself, but it logs on the SVN with the typed user name
and password. All the project update/commit actions on the repository are performed by the server on the
user’s request. Hence, if you explore the repository with SVN tools, you will see that all the changes are
attributed to the correct user.
For file:// type URLs, the pass-through authentication is not possible. Teamwork Server uses the same built-in
authentication method as that of the Native repository type, the users list with their crypted passwords is
maintained in a repository file. The server authenticates users using this file. Actions in a repository are
performed by the server on the user’s request. If the server is started as the NT service, all actions in the
repository will be attributed to the Local System user (unless a different user is specified in service settings). If
the server is started manually, all actions in the repository will be attributed to the user who started the server.
This difference can only be seen when examining the SVN repository with SVN native tools. When looking at
the project versions with the MagicDraw client, all commit actions will be attributed to the uses who performed
them.
When a project file is committed into the SVN repository, the server stores
additional auxiliary information about the project in an additional directory. For
example, if you commit the MyProject.mdzip project into the server, the
auxiliary information will be stored in the MyProject_files directory nearby. Do
not delete this directory from the repository.
For best performance, Teamwork Server and SVN repository should have a good link between them. Optimally,
Teamwork Server could run on the machine where the repository is installed.
ClearCase Repository
The teamwork Server can be configured to use ClearCase repository as a back-end. In this mode, the
Teamwork Server retrieves and commits project versions into the ClearCase repository.
To use this repository type, the ClearCase client must be correctly installed on the computer where the Server
runs. It must be able to launch cleartool executable - cleartool executable must be in PATH and have
appropriate permissions. The supported ClearCase client versions are v7.0 or later (earlier versions might
work, but were not tested).
All actions (update/commit) are performed in the ClearCase repository by the server on user request. If the
server is started as NT service, actions in the repository will be attributed to the Local System user (unless a
different user is specified in the service settings). If the server is started manually, all actions in the repository
will be attributed to the user who started the server. This difference can only be seen when examining the
ClearCase repository with the ClearCase native tools. When looking at the project versions with MagicDraw
client, all commit actions will be attributed to the users who performed them.
When a project file is committed into the ClearCase repository, the
server stores additional auxiliary information about the project in an
additional directory. If you commit the MyProject.mdzip project into
the server, the auxiliary information will be stored in the
MyProject_files directory nearby. Do not delete this directory from
the repository.
For best performance the Teamwork Server should run on the machine where the repository is installed.
• Run teamwork_server.exe in the server bin folder. The Teamwork Server startup
dialog opens.
1. Run teamwork_server.exe in the server bin folder. The Teamwork Server startup dialog opens.
2. Click the License Manager button. The Teamwork Server License Manager dialog opens.
3. On the opened dialog, click the Select License Key Files button to browse for a file with the
Teamwork Server license key.
4. In a browser, open a new license key file.
5. Click OK after you have changed the server license key.
Restart the server to apply changes.
1. Run teamwork_server.exe in the server bin folder. The Teamwork Server startup dialog opens.
2. Click the Change Server Port button and enter the new server port. This port is used when
launching the server, or when adding the Windows service.
The Teamwork Server exports remote objects through one port: RMI registry port.
1. Run teamwork_server.exe in the server bin folder. The Teamwork Server startup dialog
appears.
2. Click the Add Windows Service button.
3. After the service is added, select one of the following:
• Start this service from the Windows Services list.
• Reboot the computer and the Service will start automatically.
• To run the server, click the Start Server button.
This feature is available only on Windows operating systems.
#!/bin/bash
#
# chkconfig: - 91 60
# description: MagicDraw TeamWork Server
RETVAL=0
TEAMWORK_HOME="/var/MagicDraw_Teamwork_Server/bin"
prog="teamwork_server_nogui"
prog_stop="stop_teamwork_server"
desc="MagicDraw Teamwork Server"
args="SERVICE"
check() {
if [ -f /var/lock/$prog ]; then
if ps -p $(cat /var/lock/$prog 2>/dev/null) >/dev/null; then
return 0
fi
fi
return 3
}
status() {
check
if [ $? -eq 0 ]; then
echo $"${desc} is running..."
return 0
fi
echo $"${desc} is stopped"
return 3
}
start() {
check
if [ $? -eq 0 ]; then
echo $"${desc} is already started..."
return 2
fi
COUNT=0
while [ "$COUNT" -le 15 ] && [ -z $JAVA_PID ]
do
JAVA_PID=$(pgrep -P $SCRIPT_PID java)
let COUNT=COUNT+1
sleep 1
done
[ $RETVAL -eq 0 ] && echo $JAVA_PID >/var/lock/$prog
echo
}
stop() {
echo -n $"Shutting down $desc ($prog): "
$TEAMWORK_HOME/$prog_stop
RETVAL=$?
[ $RETVAL -eq 0 ] && rm -f /var/lock/$prog
return $RETVAL
}
case "$1" in
start)
start
RETVAL=$?
;;
stop)
stop
;;
restart)
stop
start
RETVAL=$?
;;
status)
status teamwork
RETVAL=$?
;;
*)
echo $"Usage: $0 {start|stop|restart|status}"
exit 3
esac
exit $RETVAL
This script can also be used in non-RedHat based GNU/Linux
distributions.
3. Change the value of the TEAMWORK_HOME variable according to the path of the Teamwork
Server installation bin folder.
4. Save the file and move it into the system directory “/etc/init.d”.
5. In the command line, type the following commands:
cd /etc/rc3.d
ln -s ../init.d/teamwork S99teamwork
You can also configure the service for runlevel using the
following command:
chkconfig --level 3 teamwork on
1. Stop the Teamwork Server (see "Stopping Teamwork Server" on page 15).
2. Start the Teamwork Server from the command line. Add the following argument:
"-changeKey -key:<path to the key file location>".
1. Run teamwork_server.exe in the server bin folder. The Teamwork Server startup dialog opens.
2. Click the Remove Windows Service button.
This feature is available only on Windows operating
systems.
In automatic upgrade, a new Teamwork Server version will be installed instead of current one. If you need to
keep the current server version without any changes, please upgrade the server manually. The manual
upgrade allows for installing a new server version on a new location. You will be able to import your old projects
and users, to test a functionality of a new server version, and only after that to remove a previous server
version.
Since the version of 16.9 commercial licenses are locked to the particular machine. You will be requested to
activate the license and receive the commercial license dedicated for the particular machine after the upgrade.
If you have already activated the license before upgrading with a new version, the license will be activated
automatically.
You should have valid Software Assurance to upgrade Teamwork
Server.
The easiest way to renew MagicDraw Teamwork Server is to upgrade it automatically. Using this feature the
upgrade with newest versions and service packs is done automatically.
Make sure server and the client versions are the same. Also it is
recommended to use the same JVM version for the server and client.
1. Stop Teamwork Server (see "Stopping Teamwork Server" on page 15) and close the
Administrator’s Console.
2. Deactivate the current license for the Teamwork Server.
3. For the Windows operating system, remove the Teamwork Server NT service, if it is added.
See the procedure "To remove Teamwork Server from the Windows services" on page 15.
Skip this step for other operating systems.
4. Start the Teamwork Server GUI version (see "Starting the server using GUI" on page 10).
5. When the dialog opens, click the Check for Updates button to check and download program
upgrades and updates.
If the HTTP Proxy Server Connection dialog opens, click Use
HTTP proxy server if you want to use a proxy server. Enter required
values and click OK when you are done. Checking for updates
starts.
6. The Update Information dialog opens wherein information about available updates is
displayed. Click the Update to New Version button to start upgrading the server.
7. When the automatic update finishes, the Import Configuration dialog opens. In this dialog,
click the Import button to import all teamwork projects and users from the previous Teamwork
Server version.
8. After upgrading the server to version 16.9 or later, you will be requested to apply the
commercial license dedicated for the particular machine. Please visit the https://
www.nomagic.com/support/installation-and-use/teamwork-server-install.html#activating for
instructions on how to get the Teamwork Server license.
Teamwork Server and its client must be of the same version.
Otherwise, the clients will not be able to connect to the server.
Installation of a newer version does not detect nor remove the current server version. All projects and users can
be imported to the new server.
For the description of importing procedures, see "Importing
projects and users from earlier versions of Teamwork Server" on
page 19.
The version of project server and client should be the same and cannot be mixed up. Also it is recommended to
use the same JVM version for the server and client.
• Under Choose Java Virtual Machine, click Use the Java VM installed with this
application.
5. Start newly installed Teamwork Server. The Import Configuration dialog opens.
The import time can take more than 1 hour 30 minutes. Importing
time depends on quantity and size of projects you are importing.
6. In the Teamwork Server License Manager dialog, enter the license key.
3. For the Windows operating system, remove the Teamwork Server NT service, if it was added
(see "To remove Teamwork Server from the Windows services" on page 15). Skip this step for
other operating systems.
4. Extract the MD_UML_<version number>_teamwork_server_no_installs.zip.
5. Start the new Teamwork Server without GUI (see "To start Teamwork Server from the
command line" on page 12).
There is no automatic Teamwork Server upgrade without GUI, only
manual upgrade.
C:\WINNT\Profiles\All Users\ApplicationData\.magicdrawserver\<version
number>\projects on Windows NT4
Note that the projects folder is created automatically after starting Teamwork Server for
the first time.
Note that the projects folder is created automatically after starting Teamwork Server for
the first time.
muserver.projects_directory=C\:\\ProgramData\\.magicdrawserver\\17.0.5\\projects
By changing the repository location, you can indicate the projects folder that contains the project you want to
import.
1. In the Teamwork Administrator’s Console window, Repository tab, next to the Location box,
Let’s say, we are moving Teamwork Server from the computer A to the computer B. In order to do that, follow
the instructions below.
1. Stop Teamwork Server in the computer A (see "Stopping Teamwork Server" on page 15).
2. In a computer A, back up all data in case unsuccessful data transfer.
3. Install Teamwork Server in the computer B.
4. In the computer A, copy the .magicdrawserver folder and paste them in the same location in the
computer B.
If your projects were stored other where than in the Project folder,
copy that folder wherein Teamwork Server projects were stored and
the path to that repository. In the computer B, restore the path to the
same repository and paste the folder with projects to that location.
For more information about how to change the pats to a repository,
see the procedure "To change a repository path" on page 21.
5. Start Teamwork Server (See "Managing Teamwork Server" on page 22).
Starting Teamwork Server for the first time, you will be required to
enter a license key. For more information about license activation,
see at http://www.nomagic.com/support/installation-and-use/
teamwork-server-install.html#activating.
6. You can remove back up date in the computer A (optional).
Before removing back up data from the computer A, make sure that
all data and server configurations are restored in the computer B
after moving.
2. Near to the Repository Location box, click the button, to select the repository path.
Section "Customizing Teamwork Server properties" on page 23 describes how to set general Teamwork
Server properties.
Section "Changing Administrator Password" on page 24 describes how to change the administrator’s
password.
Section "Users Management" on page 24 describes how to manage users, user permissions, and how to
assign projects to users.
Section "Starting the Administrator’s Console" on page 28 describes how to start the Administrator’s console.
Section "Administrator’s Console Dialog" on page 28 describes the Administrator’s console dialog tabs.
Section "Data Migration between Different Repositories" on page 46 describes how to migrate data between
three different repositories.
Section "Changing Teamwork Server Debugging Mode" on page 47 describes how to change the Teamwork
Server debugging mode.
For better understanding an option you can read its description that tells what is the effect of changing the
option value.
Users Management
You can create two types of users in Teamwork Server: normal users that may have different permissions and
users with administrator rights. Users with administrator rights can do the following actions in Teamwork Server:
• Manage users.
• Create projects and assign users to them.
• Manage project versions and branches.
• Set user permissions for the system and projects (the read and edit modes are set by default).
• Remove users and projects from Teamwork.
The Teamwork Server users have their own user accounts (including login names and passwords given by the
administrator) and various types of permissions. According to the storage place of the user accounts, users can
be either:
• Native - the user’s account data is stored locally.
• External - the user’s account data is stored in the external database (Subversion/ClearCase
and/or LDAP). Only the login name of an external user is stored locally.
You can create, edit, or remove both types of Teamwork users regardless of whether the integration with any
external database (Subversion, ClearCase, LDAP) is enabled or disabled. The names of native and external
users are unique per single server.
You can convert an external user to a native one and vice versa. Users with administrator rights can change a
specific user's type by editing the user's account information, or convert a whole list of active Teamwork users
by using the Teamwork Administrator's console.
You will be connected to Teamwork Server once the authorization process, which will prompt for your user ID
(login name and password), has been completed. Upon verification, you can work with the system.
If there are two MagicDraw clients with the same login name, only
one client is allowed to log into Teamwork Server at a time.
You can manage users in Teamwork Administrator’s Console or in MagicDraw UML when connection to
Teamwork Server is established.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. Click the Add button. The Add User dialog opens.
3. Enter the user’s login name, full name for better identification, and password.
4. Click OK.
5. Select the types of system permissions for the user in the Permissions list (see "User
permissions" on page 27).
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. Click the Add button. The Add User dialog opens.
3. Enter the user’s login name and full name for better identification.
4. Select the External User check box.
5. Click OK.
6. Select the types of system permissions for the user in the Permissions list.
As you cannot set a password for an external user in MagicDraw's Teamwork
Server, use an appropriate tool for managing the external database (Subversion,
ClearCase, or LDAP) wherein the user's account is stored.
To convert a native user to an external one by editing the user's account information
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. Click the Edit button. The Edit User dialog opens.
3. Enter the user’s full name.
4. Select the External User check box.
5. Click OK.
• The password of a native user, who has been converted to an external user,
will be retained. However, it will not be used in the user authentication.
• The user's native password will be enabled again only if the user is converted
back to a native user.
All converted users will be able to log into Teamwork Server only if they are
available in the external user sources (LDAP, Subversion, or ClearCase server to
which your server is integrated).
1. Start Teamwork Administrator’s Console (see "Starting the Administrator’s Console" on
page 28).
2. In the Active Users tab, click the Convert Native Users to External button (see "Active Users
tab" on page 29).
3. Click Yes to confirm your decision.
You will be informed once the conversion has been completed. A Teamwork Server's user conversion can be:
• Successful - when all the users are converted from native to external. In this case the
informational message is displayed, and you can check the list of all converted users in the
server log.
• Unsuccessful - when the conversion failed. In this case an error message is displayed, and
you can see the server log for more details.
• Non-applicable - when there are no users to convert from native to external. In this case an
informational message is displayed.
For the information about the server log file, see “Log File tab” on page 34.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. Click the Edit button. The Edit User dialog opens.
3. Enter the user’s full name.
4. Clear the External User check box.
5. Type and retype the password.
If the converted user used to be a native user, the password will be
the same one used when he or she was a native.
6. Click OK.
• All converted users will not be able to log into Teamwork Server as they do not
have passwords; therefore, the administrator has to set up a password for each
user after the conversion.
• If the converted user used to be a native user, the password will be reset to the
same one used when he or she was a native.
1. Start Teamwork Administrator’s Console (see "Starting the Administrator’s Console" on
page 28).
2. On the Active Users tab, click the Convert External Users to Native button (see "Active
Users tab" on page 29).
3. Click Yes to confirm your decision.
You will be informed once the conversion has been completed. A Teamwork Server's user conversion can be:
• Successful - when all the users are converted from external to native. In this case an
informational message is displayed, and you can check the list of all converted users in the
server log.
• Unsuccessful - when the conversion failed. In this case an error message is displayed, and
you can see the server log for more details.
• Non-applicable - when there are no users to convert from external to native. In this case an
informational message is displayed.
For the information about the server log file,
see "Log File tab" on page 34.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. In the Users area, select the user and click Remove.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. Click More, if you do not see the Teamwork projects list. The list of available Teamwork projects
is displayed in the Available Projects area.
3. Select a project you want to assign to the selected user.
4. Click the << button to move the selected project to the Assigned Projects list.
5. Click OK when you are done.
• Once a user has been added to a project, the default user rights will be created
allowing the user to access the project only according to the rights given.
• The system permissions have a higher priority over the project permissions. For
example, a user whose system permissions allow model editing can edit all
projects, even if the user does not have rights to edit the projects.
User permissions
You can give several types of permissions to the Teamwork users to coordinate the work of the whole team.
You can specify the types of user permissions in the Edit Users dialog.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. A list of users and their permissions is presented in the Permissions area.
1. From the Collaborate menu, select Users. The Edit Users dialog opens.
2. In the Users area, select a user that permissions you want to edit.
3. Select the check box to give or clear it to remove the selected permission in the Permissions
area.
1. From the Collaborate menu, select Projects. The Edit Projects dialog opens.
2. Click More, if you do not see the unassigned users list. The list of available users is displayed
in the Available Users area.
3. Select the user you want to assign to the selected project.
4. Click the << button to move the selected user to the Assigned Users list.
5. Click OK when you are done.
When a user is added to a project, default user rights are created, allowing
the user to access the project according to the rights given.
You can also start the Administrator’s Console from the client side.
The teamwork_administrator.exe file is located in <MagicDraw
installation directory>\collaboration.
In the Active Users tab, the administrator is able to observer all the users who are currently connected to the
Teamwork Server.
The GUI elements of the Active Users tab are described in the following table.
Projects tab
The Projects tab displays, names, authors, and the status of all projects that are stored on the server.
Import from and export to native repository as well as synchronization can be triggered in the Projects tab.
To synchronize projects on the current server with data from selected location
1. Click the Synchronize from Native Repository button. The Select Native Repository dialog
opens.
2. Select the directory where the data to synchronize is stored.
This must be a directory that stores project versions in the Teamwork
Server's Native repository format, for example, a directory, to which you
have exported the projects (see the procedure "To export projects from
the current server" on page 32), or a directory on which the server
operated in the past.
3. Click Synchronize.
The contents of the selected directory is imported into the current server. After the
synchronization, the server assumes that the synchronized projects are foreign. Therefore,
these projects cannot be modified, but the synchronization can be repeated to update them with
new data.
1. Click the Import from Native Repository button. The Projects Import Wizard opens.
2. Select the directory where the projects to import are stored.
This must be a directory that stores project versions in the Teamwork
Server's Native repository format, for example, a directory, to which you
have exported the projects (see the procedure "To export projects from
the current server" on page 32), or a directory on which the server
operated in the past.
3. Click Next.
To restore files from the backup that was made while upgrading the
server, in the Directory box, type the path to the backup folder.
4. In the Projects list, all the projects that are stored in the directory are displayed. Select any
projects to import from the list or click on the Select All button to import all projects. Click Next.
5. .The Resolve Naming Conflicts tab appears. The import process will check if the selected
projects/used projects/profiles already exist in the selected destination directory server
(comparison is done by name). There are two ways to resolve naming conflicts:
• Use destination. The used project will not be imported and all other projects/used
projects, which are being imported, will be modified to use the already existing used
project/profile in the current server.
• Overwrite destination. Import the selected profiles or used projects and commit
them by overwriting the existing ones in the current server.
6. Click Finish.
The contents of the selected directory is imported into the current server. After the import, the
current sever assumes the ownership of the imported projects. Therefore, they can be modified,
but the import cannot be repeated.
1. Click the Export to Native Repository button. The Project Export Wizard opens.
2. Select projects/used projects/profiles to export from the list, or click on the Select All button to
import all projects/used projects/profiles. Click Next.
3. Select the directory to export the selected project. This can be either a new directory or a
directory where the Native repository is stored.
Select a project and from the Teamwork Administrative's Console dialog and click Details.
The Project dialog opens.
If errors occur on the Teamwork Server, use the Log File tab to view the error message.
Properties tab
Since version 17.0, you can transfer data in a more secure way using the secured connection (SSL).
If the SSL connection is established in the server side, you should
also use the SSL connection in the client side when connecting to
the server.
In order to use the SSL connection, two types of certificates are needed: one for the server and one for the
client. Certificates must be in a Java Key Store format.
The server certificate is automatically placed in the <Teamwork Server installation directory>\cert folder after
the SSL configuration is done.
The client certificate should be located manually. You should create a folder named “certs” and place into it
these two files:
1. A client certificate named cert.jks.
2. A file named cert.pass wherein the certificate password is typed.
In this case applications MagicDraw and Teamwork Administrator’s Console are Teamwork server clients. Both
of it should have the client certificate. Hence the certs folder should be placed in two locations:
• <Teamwork Administrator’s Console installation directory> for the Teamwork Administrator’s
Console application
• <LOCAL_APPDATA1>\.magicdraw\<version number> on Windows OS or
<user.home>\.magicdraw\<version number> on other OS for the MagicDraw application (It can
be located in the folder <MagicDraw installation directory> either but the user home folder is the
default one)
If the Teamwork Administrator’s Console is not installed in the separate
location, its installation directory is the same as Teamwork Server installation
directory or the MagicDraw installation directory (if installed together).
You can get certificates from your system administrator or generate them by yourself.
For more information about generating certificates,
see procedure "To generate certificates" on page 37.
1. To find out the location of MagicDraw configuration files you may open Help > About MagicDraw and select the Environment tab in
the MagicDraw application. Then click the Configuration Files link.
To generate certificates
We recommend you to use the KeyTool IUI application for generating certificates.
This is a free tool that can be downloaded from the Internet.
1. Run the KeyTool IUI application.
2. Create empty files for storing certificates:
2.1. Select Create > Keystore.
2.2. Create an empty keystore file for the server. Do the following:
2.2.1 In the Keystore file dialog, set the location of the file and type a file name.
2.2.2 In the Keystore password dialog, type the password for the server keystore
file and click OK.
2.3. Create an empty keystore file for the client. Do the following:
2.3.1 In the Keystore file dialog, set the location of the file and type a file name.
For easier certificate transfer in next steps create a
new folder “certs” and save the file named cert.jks in it.
2.3.2 In the Keystore password dialog, type the password for the client keystore
file and click OK.
3. Create a RSA keypair for the server:
3.1. Select Create > Keystore’s entry > Private key, with vers. #3 > RSA.
3.2. In the Keystore file dialog, the Source section, open the created server keystore file
and type a password.
3.3. Provide the required information in the Target section and click OK.
3.4. The dialog for creating a new alias will open. Type a new private key entry’s alias name
and a password for it. Click OK.
3.5. You will see the created alias. Close the dialog.
4. Exclude a public key from the keypair to provide it to the client:
4.1. Select Export > Private’s key first certificate in chain > As simple certificate file.
4.2. In the Keystore file dialog, the Source section, open the server keystore file and type
its password.
4.3. Create a file, whereto the key will be exported. In the Certificate file dialog, the Target
section, set a location and type a file name for the client certificate. Click OK.
4.4. The dialog for selecting an alias will open. Select from the list the alias that has been
created in step 3.4 and type its password. You will be able to see the created certificate.
5. Import a public key into the client certificate:
5.1. Select Import > Keystore’s entry > Trusted certificate > Regular certificate.
5.2. In the Source section set the certificate file, which has been created in step 4, as a
regular certificate file.
5.3. In the Target section set the client keystore file client.jks as a keystore file and click OK.
5.4. The dialog will open asking to enter a new alias name.Enter the alias name created in
step 3.4 and click OK.
5.5. Some pop-up windows will open informing about the generation process. Close all of
them after reviewing.
Generated certificates are ready to use now. Paste them into the right location. For the information on how to
configure SSL in a server side please refer to procedure "To enable the secured connection (SSL)" on
page 36.
Repository tab
Repository tab is a very important tab. It contains information about the repository of the server. This tab
determines where and how the Teamwork Server stores projects and their version information, user lists, etc.
For more information about repositories,
see "Teamwork System Design" on page 7.
Restart the Server to apply changes.
There are 3 different types of repository to choose in the Repository type combo box. General information
about the types is described in the Section "Teamwork System Design" on page 7. This section, presents the
configuration fields in detail.
The Repository tab layout changes with the type of repository selected.
Native repository
The Native repository type is the simplest one to configure. There is only one editable parameter, that is the
Repository Location.
Figure 10 -- Teamwork Administrator's Console. Repository tab for Native repository type
The GUI elements of the Repository tab for the Native repository type are described in the following table.
SVN repository
Figure 11 -- Teamwork Administrator's Console. Repository tab for SVN repository type
When configuring Teamwork Server to use the SVN repository, the SVN client executable must be installed on
the computer where Teamwork Server runs. The SVN client executable must be accessible on the system's
PATH and have appropriate permissions to execute. The SVN repository itself can be located on the same or
different computer than the Teamwork Server runs.
This test ensures that the Teamwork Server is able to access and write the
cache directory, access (and log in with the specified administrator user and
password) the SVN repository in the specified location. It also checks the
configuration and system profile locations on the SVN repository (the
administrator must be able to write there).
ClearCase repository
Figure 12 -- Teamwork Administrator's Console. Repository tab for Clear Case repository type
The ClearCase type of repository has a few parameters. When configuring this repository, the ClearCase client
part must be installed on the computer where the Teamwork Server runs. The Teamwork Server will use
cleartool executable to access and manage files in the ClearCase repository, so cleartool must be accessible -
the PATH and have appropriate permissions to run. Also, the ClearCase repository must have the right
permissions for the user, the Teamwork Server.
The ability to access the server administrative functions via the command line utility facilitates the scriptable
management of Teamwork Server. This enables automation of routine administrative tasks, such as permission
management. Single command can be given through the command line utility parameters. Multiple commands
can be given in bulk through the input stream. Command results are provided through exit status codes, thus
enabling conditional script execution.
To start performing the administrative tasks using the command line utility
You are successfully logged into the server and can start performing server administration
tasks.
The following table includes the command line utility commands for performing administrative tasks and their
brief descriptions.
Command Description
mkuser Creates an external user.
rmuser Deletes the given user.
lsuser Lists all the users or tests if the given user exists.
mkproj Creates an empty project.
rmproj Deletes the given project.
lsproj Lists all the projects or tests if the given project exists.
stperm Sets the specified permission of the specified user on a given project.
clperm Clears the specified permission of the specified user on a given project.
lsperm Lists permissions of the specified user on a given project.
For detailed command descriptions, refer to the help, which is printed after typing any of the following
commands at the command line:
• teamwork_console.exe --help
• teamwork_console.exe -h
Any of the commands in the preceding table can be typed at the command line immediately after the
servername parameter.
teamwork_console.exe -u Administrator -p Administrator localhost lsuser
If your Teamwork Server runs on the Linux or other Unix-based operating system,
you can escape typing a long command every time by using once the following
command:
alias tc="./teamwork_console -u username -p password servername"
In the sequel, you can type “tc” instead of “teamwork_console -u username -p
password servername”, for example, “tc lsuser”.
The commands for performing the supported administrative tasks can be given on the input stream. This is
indicated by specifying the “-” parameter immediately after the servername parameter.
teamwork_console.exe -u Administrator -p Administrator localhost -
In this case the command line utility connects to the specified server with the given credentials, performs the
commands and exits.
This is a shell script for creating a project and assigning users to it:
#!/bin/bash
#shorthand
TC='/MagicDraw_TeamworkServer_installation_directory/bin/
teamwork_console -u Administrator -p Administrator localhost'
There are many different but similar use cases, such as migrating the data from the previous Native type
repository into a new SVN/ClearCase based repository.
It is possible for the Teamwork Server to import and export projects. This function is triggered from the
Administrators console (see section “Administrator’s Console Dialog” on page 28). Import & Export is only
possible in the Native repository type format. The Native repository type is a kind of intermediate form for
information interchange.
Note that words “import” and “export” here are used relatively to the currently running server (the one to which
administrator's console is attached):
• “import” means importing data into the current server from the designated directory;
• “export” means exporting data from the current server into the designated directory.
Migrating the server from the Native repository to the SVN/ClearCase repository
2. Open the Administrator's Console. Reconfigure the server to work with the SVN/ClearCase
repository.
3. In the Administrator’s Console, reconfigure the server to work with the ClearCase/SVN
repository.
4. Restart server to use the new repository. The Server starts. There is nothing in the repository
but the profiles needed to work with MagicDraw.
5. Login again to the Administrator’s Console and trigger project import. Select the directory to
import from, the same directory from where the export was performed.
6. Projects are now in a new SVN/ClearCase repository.
LDAP Support
LDAP Integration allows Teamwork Server to authenticate its users against LDAP servers. LDAP Integration
enables pass-through MagicDraw authentication against LDAP servers, by passing client's authentication
information to LDAP servers.
LDAP Integration supports Simple User+Password and SASL authentication, SSL/TLS protocols, and several
LDAP servers configured for a single integration.
5. Click Apply Changes when you are done. You will be required to test the connection against
the LDAP server.
For more information about the connection testing,
see "Connection Testing" on page 56.
Connection Settings
Connection Settings specifies network and security settings for connecting to LDAP servers.
Authentication Settings
LDAP Integration supports two most popular LDAP authentication methods. They are as follows:
• Simple User+Password (see "Authentication settings for the Simple User+Password
authentication type" on page 51).
• SASL (see "Authentication settings for the SASL authentication type" on page 54)
Teamwork Server transforms user credentials entered in MagicDraw to LDAP authentication credentials by
using the templates in authentication settings. After successful authentication to LDAP, a special user for each
authenticated LDAP user is created in Teamwork Server. They differ from ordinary users as they have no
passwords (in order to complete authentication, authentication to LDAP server(s) is used). You can perform
various actions for these users, otherwise. It is possible to setup permissions, remove users, and do other
common actions with users.
You can automatically create a proxy user account for Teamwork Server. Select Auto-add unknown users if
they login successfully checkbox and authenticated user credentials without password will be stored in
Teamwork Server on successful login.
Figure 15 -- Teamwork Administrator’s Console, LDAP Integration tab. Selecting Authentication Type
Using the Simple User+Password authentication type, you can select the following options:
• Use User DN template
• Retrieve User DN by using an LDAP query
Figure 16 -- Teamwork Administrator’s Console, LDAP Integration tab. Authentication Settings (Simple User+Password)
Figure 17 -- Teamwork Administrator’s Console, LDAP Integration tab. Authentication Settings (SASL)
User DN is retrieved in the same way as it is done if the Simple User+Password authentication type is
enabled (either by using a static User DN template or by querying the LDAP server(s) for User DN). When the
user logs in to the LDAP server, this connection is further reused for retrieving user information.
If the user information retrieval is disabled or User DN attributes are not accessible to the authenticated user,
Teamwork Server creates an external user with the login name that was specified by the user on the
authentication.
Figure 18 -- Teamwork Administrator’s Console, LDAP Integration tab. User Data Retrieval Settings
This will form the Name of the created user out of two LDAP attributes - cn and
sn.
Settings that are active when the Use User DN templaten is selected:
User DN User DN specifies a template used to search for specific DN by the supplied
login name. An example:
cn=$(login), dc=example, dc=com
Settings that are active when the Retrieve User DN by using an LDAP query is selected:
Query An LDAP query for retrieving User DN, for example:
uid=$(login)
Search Base DN from which the searching should begin, for example:
dc=example,dc=com
Search Scope Specifies, whether the search must be restricted to the directly owned DNs
only or it must be performed in the whole subtree.
Choose one of the following options:
• One level
• Subtree
Connection Testing
After you have specified the LDAP Integration setting values, you can test the connection to LDAP server.
1. In the LDAP Integration tab of Teamwork Administrator’s Console, click Test Connection.
The Test Connection dialog opens.
2. Type user’s login name, password, and click OK.
3. The message with connection results appears.
When Teamwork is integrated with Subversion only, client's authentication information is passed to the
Subversion server. When Teamwork Server is integrated with Subversion and LDAP, client's authentication
information is passed for both Subversion server(s) and LDAP server(s), but only successful authentication to
LDAP server(s) successfully logs the user into Teamwork Server.
1. Export the CA and AD server certificates to the DER encoded binary.cer files using the
Microsoft Management Console, the Certificates Snap-in.
Do not include private keys while exporting.
2. Import the CA and AD server certificates (.cer files) to Java KeyStore (.jks file), using the
KeyTool IUI. Do the following steps:
2.1. Run the KeyTool IUI.
2.2. Double-click Create on the tree, and then click Keystore to create a new keystore file.
2.3. Choose the JKS format and save a new keystore file.
Figure 19 -- Creating a new keystore file with the KeyTool IUI (steps 2.2, 2.3, and 2.4)
2.5. On the tree, double-click Import, Keystore's Entry, and Trusted Certificate, and then
click Regular Certificate to import the .cer files into Java KeyStore.
2.6. Select the created keystore file as the target.
2.7. Select the exported CA certificate file (.cer) as the source. Enter the keystore password
and click OK. Enter “CAAlias” as the alias for the CA certificate and click OK.
2.8. Select the exported AD server certificate file (.cer) as the source. Enter the keystore
password and click OK. Enter the full name of the AD server as the alias for the AD
certificate and click OK.
Figure 20 -- Importing certificate files into the keystore file (steps 2.5, 2.6, 2.7, and 2.8)
The subsequent steps for the Teamwork Server integration with SSL-enabled Active Directory are the same as
for the integration with any other LDAP server. This procedure is described in section “Enabling LDAP
Integration” on page 48.
4. Create a local group for SSH tunnel users. To do this right-click on My Computer and then
select Manage. In the Local Users and Groups section right-click on Group and choose New
Group. The New Group dialog opens (see Figure 22 on page 61).
4.1. Enter a new groupname, for example, SSH.
4.2. Add the “tunnel” user to the SSH group.
5. Create SSH-aware local password file with 'tunnel' user entry. Any users in this password file
will be able to log on with SSH. To create SSH-aware local password file run command prompt
(click "Start"-> "Run", then type "cmd" and click Enter) and then type the following commands:
cd C:\Program Files\OpenSSH\bin
mkgroup -l >> ..\etc\group
mkpasswd -l -u tunnel >> ..\etc\passwd
6. Start OpenSSH Server service from your control panel. To do this right-click on My Computer
and then select Manage. In the Services and Application section, the Services item right-
click on the OpenSSH Server service and choose Start.
7. Test the SSH server.
7.1. Type “ssh tunnel@localhost” from your command prompt.
In the command prompt go to the C:\Program Files\OpenSSH\bin and type the 'netstat -na' command. You will
get the list of all connections. The state of the port should be "LISTENING" while the SSH server is running.
The SSH client package is installed to "C:\Program Files\OpenSSH" by default and added into your PATH
variable.
2. Establish an SSH tunnel by logging into Teamwork server SSH service from the command
prompt:
For example, the following command will establish an SSH encrypted tunnel from client port 1100 to server port
1100. When connecting to localhost:1100, the packets get encrypted and sent to twserver:1100, where actual
Teamwork server resides:
ssh -L 1100:twserver:1100 tunnel@twserver
Teamwork server port is the port the Teamwork server is running (usually 1100)
You are logged as user "tunnel" to the SSH service on Teamwork server machine. Leave the session opened,
as killing it also kills the SSH tunnel used for the MagicDraw Client.
3. Open MagicDraw Client. Use localhost:localport when connecting to the Teamwork server.
Localport is 1100 in this case as we used it when created the tunnel.
Using any other value than "localhost" or "127.0.0.1" will fail, even if
connected to the actual machine name, resolved by DNS. This is
because tunnel starts on loopback interface of your workstation only
for security reason.
Synchronization Overview
Ability to synchronize data between several Teamwork Servers is available as of version 17.0.3. One Teamwork
Server can now expose and make accessible the projects of the different servers together with owned projects.
Server users can access remote projects (coming from the different locations) in a read-only mode and can
perform any read-only actions.
There can be more than two servers in the pool. Each server can separately synchronize data from many other
servers.
The main usage scenario for the inter-server synchronization capability is multisite Teamwork Server
deployment.
Very often a large company has several different sites (often in different countries) where projects are
developed. Usually the site has very good internal connectivity - LAN-class speeds (1Gbit/s, or at least
multiMbit/s) are available inside the site; while connectivity between the sites is not so good - WAN-class
speeds (several Mbit/s) are available between sites.
There are cases when developers at one site are responsible for development of some subset of projects, while
developers at other sites just need to use (but not modify) the artifacts produced by the first team.
In the earlier versions, company had to choose one site where a single Teamwork Server will be deployed and
set up the remote connections for the developers joining from the other sites.The server performance (project
open, commit, update, element locking times) for developers of that site were good. But the developers at other
sites suffered from degraded performance due to reduced connectivity parameters. This is especially important
when managing large projects.
As of this version, company can deploy many Teamwork Servers - one per each site. Developers of that site
can connect to their local server and work at LAN-speeds. Synchronization between servers can then be set
up.
With the synchronization set up, developers can access the projects of other teams on their own local server.
Any action that is not changing the project is acceptable - users can open the remote projects, browse, search,
analyze then, generate reports and most importantly - include remote projects as used projects/libraries into
their own projects.
For example, team at one site is responsible for developing requirements project; team at another site can take
this requirements project and incorporate them into their implementation project(s).
Please note that scenario where developers at different sites edit the same project is not covered. In this case
one server will be designated as the home server of the project, and any developer, who needs to edit it, has to
log onto the home server of the project to edit it.
There are additional, less frequent scenarios when synchronization between the servers can become handy.
For example there is a scenario: there is a project set which has to be readable by a huge number of users, but
is developed only by the small number of developers. In this case company may want to set up a small main
development server where developers can work unaffected by large numbers of browsing users and a large
server for handling all read-only traffic (possibly in a different network security zone etc).
The licensing scheme is very simple. Each Teamwork Server deployment is licensed separately - multiple sites
synchronizing between each other is NOT counted as one big deployment, but as many separate deployments.
There are no additional charges (such as a separate "enterprise" edition of the server) for using
synchronization feature.
Characteristics
Using Synchronization
As described earlier, synchronization process is launched between two given servers - source (providing the
information) and target (receiving the information). Target server is the driving server - it actively connects to the
source server and pulls the necessary data. Synchronization can be bidirectional, but for each direction a
separate synchronization must be launched. Both directions are independent of each other and can be present
or absent as necessary. Several synchronization processes can be run in parallel on the same target server,
pulling the data from several different source servers.
Synchronization process takes the project history data (including the project and category names, all the
commits, comments, and branches) from the source server and uploads it to the target server. This process is
incremental, so only the changes from last synchronization are transferred. Target server users can then
access this data in the same manner as local projects of the target. These projects are listed among other
projects, can be opened, or attached as used projects/libraries to other projects. The only limitation is that they
cannot be edited and no further versions can be committed. Project can only be edited on the home
server.Project permissions of the source server are not synchronized - target server administrator can specify
the different access rules to the remote projects. Note that specifying write permissions on remote project has
no effect.
The remote project can be removed from the target server, but please note that it will reappear after the next
synchronization, unless it is also excluded from the synchronized project set of the source server. The rename
operations are handled in the same way - it is possible to rename remote project on the server, but the name
will be re-synchronized to the name in the source during the next synchronization.
Synchronization can be a long-running process, but it is not intrusive. The end users of the source and target
servers do not feel it's effects (no need to stop or restart servers during or after synchronization) except for the
possible slight performance degradation.
The results of the pending synchronization appear on the target server gradually during the course of
synchronization. So it is possible to see particular remote project X having 100 versions, then 110 after some
time and 117 after synchronization is complete. But this should not cause the problems for the clients on the
target - partial version data will never be offered (the situation where there are 116 versions of the remote
project X declared, but the 116th version is only half-way downloaded are never exposed).
The remote and local projects on the server are almost indistinguishable. To differentiate between the two
kinds, synchronization process can prepend the arbitrary prefix to the remote category names.
During synchronization, administrator provides (optionally) the prefix string and synchronization process
modifies the names of the categories, coming from the remote server. When synchronizing from multiple
source servers, each different source can have a different prefix.
For example, in the target server of GB site, projects coming from the server of American site can have “USA_”
prefix added to their category names; and projects of German site can have “DE_” prefix. So if there are
projects under category “Requirements” on the German site, on GB site these projects will be located under
“DE_Requirements”.
It is possible to configure which projects are synchronized - select all or only particular subset of the source
server projects to be synchronized to the target.
The target server specifies user/password of the source server when target server initiates the synchronization
process. Synchronization process then takes all the accessible projects from the source server. Source server
administrator can control which projects are accessible managing user permissions on his server. If the
specified synchronization user has a permission to read all the projects on the source server, then all the
projects will be synchronized during the process. If the specified user only has permissions to access particular
subset of projects, only these projects shall be synchronized during the process.
It is highly advisable to create a special dedicated user on the source
server for synchronization purposes and not reuse the real existing user
accounts. Only read permissions should be given to this account.
Note that different synchronized project sets can be specified on the source for each different target - just use
different users for each target and define the project set as described above.
Not every user can trigger synchronization procedure (on the target server). The user must have at least the
following system-wide permissions:
• Create project/category
• Edit project properties
• Rename category
It is not enough to have project-specific permissions.
During synchronization there can be situations when project was previously located in the source server but is
no longer available. Project may have been deleted in the source server or no longer accessible (by decision of
source server administrator).
The synchronization process does not and cannot delete these unavailable remote projects from the target
server automatically, because there may be other, local, projects still using this remote project. Missing remote
project is not deleted from the target server, but moved into a special category - ATTIC.
Target server administrator should inspect this category from time to time and if there are any projects, inspect
whether they are used in the local projects (perhaps using the Project Usage Map.
For more information see “Project Usage Map” in MagicDraw
UserManual.pdf ) and remove unnecessary projects.
Since connection between the source and target servers is usually goring through the WAN, it is advisable to
secure this connection - so that 3rd parties cannot access the project data in transit.
One option is to set up the SSL connection. In this case the source server needs to have public/private key pair
configured in a keystore. Target server needs to have the public key (in certificate form) of each source server it
wants to connect to.
For more information, see "Secured Connection tab" on page 35.
Another option is to set up the VPN connection between the servers so that both servers are in the same virtual
subnet. This solution is server-neutral. Nothing needs to be done on the server(s).
Running Synchronization
Synchronization procedure can be started in two different ways - using a build in scheduler in the Teamwork
and using the command line utility. As stated earlier, synchronization procedure has to be triggered separately
for each synchronization direction in case of bi-directional setup. The synchronization target server (the server
receiving the data) is the active server - the one which drives the synchronization process. When started,
synchronization process runs till completion or till the unrecoverable error appears (such as network problems)
or until interrupted by the user (when using command line utility). The synchronization progress and outcome
can be monitored in the server log.
Teamwork Server has a build in mechanism to trigger tasks at various periodicities and with various
parameters.
The tasks are specified by *.properties files located in the schedule subdirectory in the Teamwork Server install
directory. Each *.properties file describes one task, to be run at the defined periodicity. There is an example file
- synchronization.properties.example, which can be examined and copied/renamed to produce the
synchronization task(s). For example two files - syncDE_GB.properties and syncDE_US.properties can be
create to describe two synchronizations - one for synchronization from remote server on GB site, another - for
synchronization from remote server on US site.
Teamwork Server constantly monitors the schedule subdirectory for any changes in the *.properties files and
any changes take effect immediately. All other files with different endings - such as aforementioned
synchronization.properties.example - are ignored.
When scheduled tasks are run, they are run on behalf of the system user of the Teamwork Server (by the way -
this is the reason why you do not need to specify the user on the target server when running synchronization
using scheduler, but you need to specify user on target server when running synchronization from command
line utility)
Once triggered by the scheduler, synchronization runs to completion (or to unrecoverable error). The progress
of synchronization can be monitored in the server log file. There is no facility to stop these jobs by user
command (except for full server stop).
If the synchronization job takes longer than the schedule period, and the next schedule event happens before
previous job has finished, the event is suppressed. The second synchronization is NOT started in parallel; the
running synchronization is allowed to continue and finish normally.
If running the synchronization through the scheduler mechanism is not flexible enough (perhaps you want to
use the mechanisms of your OS for scheduling tasks, or perhaps you want to run synchronization manually, on
demand), there is a separate command line utility for triggering the synchronization process.
The utility is located in the bin subdirectory in the Teamwork Server installation directory. The utility is called
synchronize.exe for Windows deployments and synchronize for Unix deployments. Utility has an
accompanying synchronize.properties file.
This utility can be run from any computer (not just the servers), but most frequently it will be run on the target
server. When utility is run, it connects to the target server and initiates the synchronization procedure on it.
There is no additional load on the computer which runs the utility - procedure is run on the server side.
Utility exits when the synchronization process is complete. It provides the exit codes - 0 in case synchronization
is successful and >0 in case of errors. This exit code can then be used for shell scripting purposes (such as
restarting synchronization in case of failure or similar).
Please note that since the synchronization procedure is running on the server, exiting (or crashing, or killing)
the utility will not stop the synchronization process on the server. But the utility does accept the Ctrl+C keyboard
signal and stops the synchronization process on the server cleanly.
Utility takes 6 or 7 parameters (the last one - prefix - is optional). Parameters are mostly the same as the
scheduled job, but there are additional parameters for specifying the target server to connect to (since the
connection layout is now utility =>target server =>source server).
<prefix>
The categories from the target server, which are pulled from the source server, will be prefixed
with the specified prefix. This allows better differentiation of projects on the target - it is easier to
see which projects are remote and which projects are local.
Options:
-h, --help
This help message.
-q, --quiet
Quiet synchronization. No information is displayed on system output stream during
synchronization. Errors are still reported on system error stream.
--ssl
Implies --targetssl and --sourcessl.
--targetssl
Use SSL connection between utility and target server.
--sourcessl
Use SSL connection between target server and source.
Example
synchronize tws.company.de sysop p@Ssw0rd tws.company.ca:1102 sync ssApcnys CA_
This command connects to the tws.company.de server with sysop user, p@Ssw0rd password and initiates the
synchronization procedure - pulling the information from the tws.company.ca:1102 server:port, using sync user,
ssApcnys password. Categories form .ca server will get the CA_ prefix on the .de server. E.g. the category
“Requirements” from the ca server will be named "CA_Requirements" on the .de server.
If necessary, you can make the copies of utility with the pre-defined parameters. To do that, you can copy the
synchronize.exe and synchronize.properties to files with different names and modify the APP_ARGS
parameter inside the copied *.properties file - specifying the necessary arguments.
For example, you can copy synchronize.exe to synchronizeDE_GB.exe and copy synchronize.properties to
synchronizeDE_GB.properties. Inside the synchronizeDE_GB.properties, you can specify DE and GB server
parameters in the APP_ARGS line in the same way as you would specify parameters on the command line.
After this you get a single executable (with all parameters pre-specified), that can be run by one click.
MagicDraw Teamwork Server 18.1 supports transferring project data from one Teamwork Server to another by
using any external storage device, such as CD, DVD, hard disc, or flash memory device. The updated version
of the shared project can be transferred back to the sharing server and smoothly merged with the original
project version. Furthermore, the same project version can be given to several contributors simultaneously, and
the contributions to the model they make can be successfully merged as well.
To manage the data transfer between isolated servers, exploit the functionality of the Projects tab in the
Teamwork Administrator's Console.
Here is the workflow on how to interchange projects between two isolated servers: