Virtual DataPort Server Configuration¶
To begin using the VCS integration, configure the VCS settings of the Virtual DataPort server. To do this, click VCS management on the menu Administration and then, click Configuration.
Configure the following parameters:
Use version control. Select it to enable the Version Control System support.
After this, you have to enable it in the databases where you want to use the VCS integration (see section Environment Management). By default, it is disabled for all the databases.
Environment. Shows the selected environment. The first time you configure the VCS integration, there is no selected environment.
Version control system. Default version control system.
URL. The path to the root of the version control repository.
Example for GIT:
Example for Subversion:
For Microsoft TFS:
There are two different parts in this TFS URL:
http://acme:8080/tfs/collection1is the URL of the TFS server.
project1/folderA/branch1is used to obtain an appropriate “working folder” (a “working folder” is a TFS concept akin to the Subversion “working copy” concept)
Although it is not mandatory, we recommend adding a
<branch>to the URL, as it will be easier to add development branches in the future if we begin to work with branches from the start.
Use default branch (only for Git): if selected, the “default branch” of the VCS system of Virtual DataPort will be the default branch of this repository. If cleared, enter the branch you want to use by default (the branch needs to exist in the repository).
The VCS system uses this “default branch” for two things:
When you import a new database from a Git repository, the VCS system will list the Denodo databases stored in this branch. Once you import a database, you can change its branch.
When you enable VCS on an existing database (that is, a database that already exists on this server but it is not synchronized with a Git repository). The dialog to enable VCS on a database has a field to enter the Git branch associated with this database. The value of this field is the default branch but you can change it.
Use user and password. If cleared, Virtual DataPort will use the authentication methods configured in the host operating system.
When working with Microsoft TFS, if you clear this check box, Virtual DataPort will use the Windows authentication to connect to the TFS server. Therefore, the Windows user that launched Virtual DataPort must have granted the necessary privileges.
Follow these steps to connect to Git using SSH authentication:
Clear the check box Use user and password.
Log in to the computer where Virtual DataPort runs, with the same user account you use to start Virtual DataPort.
In this computer, check that the Git server is in the “known hosts” file of this computer (usually located in
$HOME/.ssh/known_hostson Linux and
If this file does not exist or does not contain an entry for the Git server, run this command:
ssh -o HostKeyAlgorithms=ssh-rsa,ssh-dss <host name of the Git server>
Execute the command below to generate an SSH key pair. They must be RSA keys and they cannot have passphrase.
ssh-keygen -t rsa -C "<firstname.lastname@example.org>" -b 4096 -m PEM
This generates two files in the
.sshdirectory of your home directory:
id_rsa(the private key) and
id_rsa.pub(the public key).
Upload the public key (
id_rsa.pub) to the Git server. The administrator of this Git repository will be able to help you to do this.
User. The user to access the specified repository.
When working with TFS, the user name must be entered with the format
Password. The configured user’s password.
Local repositories home. Directory where the Server will create the local repositories for the databases with VCS enabled.
Change this if you are going to work with Microsoft TFS. Microsoft TFS limits the maximum length on file paths so we strongly recommend changing this path to the shortest path possible.
C:/tfs_repo. You do not need to change this when working with GIT or Subversion.
After disabling and enabling the VCS integration or changing the “URL” of the VCS server, all the VCS metadata will be lost. That is, you will have to check in again all the elements of the databases you want to version.
When using the support for GIT, the pull operations are performed with a fetch followed by a rebase.