Upgrade with Docker

To upgrade an existing Docker installation

1. Create a backup of the Hub database while your existing Hub installation is running.

2. Stop the docker container.

3. Pull an image of the latest Hub version from the Docker Hub Repository:

   docker pull jetbrains/hub:<version>

   Where <version> is a full version number of an Hub build. See the complete list of available versions.

4. Execute the following command to run a container with the new Hub version, that you have pulled from the repository. All other attributes of the command (volumes and port mappings) must be the same as those that you used with the previous version of the Hub image:

   docker run --rm -it \ -v <path to data directory>:/opt/hub/data \ -v <path to conf directory>:/opt/hub/conf \ -v <path to logs directory>:/opt/hub/logs \ -v <path to backups directory>:/opt/hub/backups \ -p <port on host>:8080 \ jetbrains/hub:<new version> configure -J-Ddisable.configuration.wizard.on.upgrade=true

   docker run --rm -it \ -v <path to data directory>:/opt/hub/data ^ -v <path to conf directory>:/opt/hub/conf ^ -v <path to logs directory>:/opt/hub/logs ^ -v <path to backups directory>:/opt/hub/backups ^ -p <port on host>:8080 ^ jetbrains/hub:<new version> configure -J-Ddisable.configuration.wizard.on.upgrade=true

   Use the following information to customize this command as needed:

   • --rm

   • -it — a command-line flag that attaches both the input and output of the container to your terminal. This lets you see the output produced by the container as if it were running directly in your terminal.

     When the container starts, you will see the log messages generated during the initialization process as well.

     Pressing Ctrl+ C while the terminal is attached to a container's output sends an interrupt signal to the container's running process. This is similar to how you would stop a regular process in your terminal. This command causes the container to shut down, so be cautious when using Ctrl+ C in this context.

     If you want to detach the terminal from the container output without interrupting the processes running in it, press Ctrl+ P followed by Ctrl+ Q.

     For more details, please refer to the the official docker documentation.

   • -v <path to <name> directory>:/opt/hub/<name> — these options specify the location of the storage provisioned on the host machine and maps them to the corresponding /opt/hub/<name> directories inside the container.

   • -p <port on host>:8080 — parameter that defines the port mapping. This tells the host machine to listen for traffic on the <port on host> port and propagate all traffic to port 8080 inside the container.

     The listen port must be accessible to end users or to the reverse proxy server, if one is in use. This may require explicitly opening access to this port through the firewall.

   • configure — passes the specified configuration parameter to the application at startup. In this case, the parameter -J-Ddisable.configuration.wizard.on.upgrade=true tells Hub not to launch the web-based Configuration Wizard at startup. This causes the upgraded service to launch using the same configuration that was applied to the previous version.

1. Create a backup of the Hub database while your existing Hub installation is running.

2. Stop the Hub service.

3. Pull an image of the latest Hub version from the Docker Hub Repository:

   docker pull jetbrains/hub:<version>

   Where <version> is a full version number of an Hub build.

4. Create and configure Hub-specific directories on the host machine.

5. Copy the backup file of the Hub database into the backups directory, that you created in the previous step.

6. Execute the following command to run a container with Hub server and map Hub data volumes and port:

   docker run -it --name <hub-server-instance> \ -v <path to data directory>:/opt/hub/data \ -v <path to conf directory>:/opt/hub/conf \ -v <path to logs directory>:/opt/hub/logs \ -v <path to backups directory>:/opt/hub/backups \ -p <port on host>:8080 \ jetbrains/hub:<version>

   docker run -it --name <hub-server-instance> ^ -v <path to data directory>:/opt/hub/data ^ -v <path to conf directory>:/opt/hub/conf ^ -v <path to logs directory>:/opt/hub/logs ^ -v <path to backups directory>:/opt/hub/backups ^ -p <port on host>:8080 ^ jetbrains/hub:<version>

   Use the following information to customize this command as needed:

   • -it — a command-line flag that attaches both the input and output of the container to your terminal. This lets you see the output produced by the container as if it were running directly in your terminal.

     When the container starts, you will see the log messages generated during the initialization process as well.

     Pressing Ctrl+ C while the terminal is attached to a container's output sends an interrupt signal to the container's running process. This is similar to how you would stop a regular process in your terminal. This command causes the container to shut down, so be cautious when using Ctrl+ C in this context.

     If you want to detach the terminal from the container output without interrupting the processes running in it, press Ctrl+ P followed by Ctrl+ Q.

     For more details, please refer to the the official docker documentation.

   • --name hub-server-instance — this is an arbitrary name for the container.

     By default, Docker generates random and sometimes cryptic names for containers. Use the --name option to give your container a more meaningful and human-readable name that makes it easier to identify and manage.

   • -v <path to <name> directory>:/opt/hub/<name> — these options specify the location of the storage provisioned on the host machine and maps them to the corresponding /opt/hub/<name> directories inside the container.

   • -p <port on host>:8080 — parameter that defines the port mapping. This tells the host machine to listen for traffic on the <port on host> port and propagate all traffic to port 8080 inside the container.

   The Hub service starts on {0.0.0.0:8080} inside the Docker container, and port 8080 is mapped to the specified <port on host>. This way the service can be accessed from any machine that has network access to the <port on host> of your host machine. For instance, if the fully qualified domain name (FQDN) of the host machine is host.mydomain.com and the <port on host> is 7777, the service will be available at http://host.mydomain.com:7777.

   The Hub service starts with a web-based Configuration Wizard. The wizard is accessible from the web address where the service is hosted.

   The Configuration Wizard is secured with a one-time access token, which you need to copy from the log output from Docker. For example:

   JetBrains Hub 2023.1 Configuration Wizard will listen inside container on {0.0.0.0:8080}/ after start and can be accessed by URL [http://<put-your-docker-HOST-name-here>:<put-host-port-mapped-to-container-port-8080-here>//?wizard_token=FOT8QIzHuktzvxtiT1Ax]

   Using the example from the previous section, the Configuration Wizard would be available from http://host.mydomain.com:7777//?wizard_token=FOT8QIzHuktzvxtiT1Ax

7. Open the URL for the Configuration Wizard in your web browser.

8. In the Configuration Wizard, click Upgrade.

   

9. On the Select Upgrade Source page, click the Select button and select the backup file as an upgrade source.

   

   Click Next.

10. On the Confirm Settings page, review the system settings.

    

11. When done, click Upgrade.

    • Hub launches its components.

      Do not close the page in the browser until the setup is complete. When the Hub server is ready, you are redirected to the login page.

12. Enter the credentials for the Hub administrator account and click the Log in button.

    • Hub opens to the Users page.

    Your Hub installation is upgraded and ready to use.