How-to Back up workstation computers with Duplicati
The following describes how to install Duplicati on various operating systems as well as how to run the software for the first time. You will also be shown how to save and restore your data.
Table of Contents:
- Installation on Windows
- Installation on Mac OS
- Installation on Linux (Ubuntu)
- Installation on Linux (Fedora)
- Running Duplicati for the first time
- Saving data
- Restoring data
Installing Duplicati
Download the most up-to-date version directly from the developers' website, which is linked above. Here you will find the download files for the most common desktop operating systems. Generally, the system you are using will be detected, and the link to the appropriate installation file will be displayed at the very top:
Installation on Windows
Start the installation by double-clicking on the downloaded .msi file. Continue by clicking Next.
Accept the End-User License Agreement and click Next.
You do not need to make any changes in this window; continue with Next.
Click on Install to start the installation process.
Complete the installation by clicking "Finish". Duplicati can be started immediately.
Duplicati has now been installed. This services runs in the background so that the backups are performed automatically on a schedule without Duplicating needing to be open in a separate window. Duplicati can be operated via a web browser. A local webserver will be installed for this purpose. The operation of Duplicati is largely identical on all platforms and will therefore be described once below. Duplicati can also be controlled remotely from another computer if this function has been activated in the settings. This may be helpful for administrators of decentralized workstations.
Installation on Mac OS
Installation starts by double-clicking on the .dmg file that was downloaded from the website linked above. In the window that appears, simply drag the Duplicati.app to the Applications folder. Duplicati will now appear in the Programs folder and can be started by double-clicking the icon.
When opening Duplicati for the first time, you will be asked whether the “Mono-Framework” should be installed. Click on Download as this framework is crucial for using Duplicati with Mac OS.
You will be redirected to a website where you can download the Mono Framework for Mac OS. Click on the relevant download link.
Double-click the downloaded .pkg file to execute it and follow the installation instructions. Special settings are not required. Background information: Duplicati was developed on Windows using the .NET Framework. The Mono Framework provides NET functionality for Linux and Mac OS.
Duplicati has now been installed. This services runs in the background so that the backups are performed automatically on a schedule without Duplicating needing to be open in a separate window. Duplicati can be operated via a web browser. A local webserver will be installed for this purpose. The operation of Duplicati is largely identical on all platforms and will therefore be described once below. Duplicati can also be controlled remotely from another computer if this function has been activated in the settings. This may be helpful for administrators of decentralized workstations.
Installation on Linux (Ubuntu)
A .deb file can be downloaded from the website linked above. Some browsers directly execute the downloaded file. Should this function not be available, start the installation by double-clicking on the .deb file, which you can find in the download folder using the file manager. In the window that appears, you can select how you want to proceed with the file.
Installation will begin when you click on Install in the next window.
Duplicati has now been installed. This services runs in the background so that the backups are performed automatically on a schedule without Duplicating needing to be open in a separate window. Duplicati can be operated via a web browser. A local webserver will be installed for this purpose. The operation of Duplicati is largely identical on all platforms and will therefore be described once below. Duplicati can also be controlled remotely from another computer if this function has been activated in the settings. This may be helpful for administrators of decentralized workstations.
Installation on Linux (Fedora)
A .rpm file can be downloaded from the website linked above. Some browsers directly execute the downloaded file. Should this function not be available, start the installation by double-clicking on the .rpm file, which you can find in the download folder using the file manager. In the window that appears, you select how you want to proceed with the file.
Installation will begin when you click on Install in the next window.
Duplicati has now been installed. This services runs in the background so that the backups are performed automatically on a schedule without Duplicating needing to be open in a separate window. Duplicati can be operated via a web browser. A local webserver will be installed for this purpose. The operation of Duplicati is largely identical on all platforms and will therefore be described once below. Duplicati can also be controlled remotely from another computer if this function has been activated in the settings. This may be helpful for administrators of decentralized workstations.
Starting Duplicati after installation (all operating systems)
Always start Duplicati through the start menu, i.e. the Programs folder of your operating system.
Duplicati communicates with the central backup storage through the standard browser in a new tab, which means that it can be operated virtually independently of the operating system being used. If necessary, you can use the web browser in the foreground. When you start Duplicati for the first time after installation, you will be asked the following question: If your machine is in a multi-user environment (i.e. the machine has more than one account), you need to set a password to prevent other users from accessing data on your account. Do you want to set a password now?
Generally, the user uses a single workstation and can click on “No, my machine has only a single account”. If this is not the case, click “Yes” and you will be asked to set a password for the Duplicati web interface. Background: Duplicati runs as a service. Therefore, users of the same computer could gain access to other users' back-up data via the Duplicati web interface. If your computer is connected to the Windows domain, please note that other users can log on to your computer.
After selecting one of the options above, you will be directed to the Duplicati home page.
Backing up data - adding a backup
The Add Backup command found in the function overview on the left allows you to create one or more backup jobs. These backup jobs differ in terms of the selection of files to be saved, the time and the frequency of the automatic backups, the potential encryption of the data, the backup destination, as well as other configuration options. A separate configuration file will be created on the local computer for each created job. These files can be exported as backups or be transferred to another computer. For faster access to the files stored in the backup destination, a database is also created on the local computer for each job (SQLite).
When you click the Add backup option, a dialog window for creating a new backup job will open.
The Configure a new backup option is preselected. The Import from a file option enables you to access exported configuration files mentioned above. After you click Next, you can adjust all the necessary settings for the backup in the following five steps. Each step will appear on a new page. You can go back to the previous steps at any time.
Step 1: General backup settings
Name: Serves to distinguish between several backup jobs
Description: Optional
Encryption: This is strongly recommended. The data is kept in an encrypted form on the server storage and cannot be viewed even by an administrator.
Passphrase: In the following two fields, you can select a passphrase with which to encrypt your data. Please be aware: This passphrase will only be stored on the local computer and in any exported configuration files. The passphrase will not be stored on the server. The backed data cannot be accessed without the passphrase. The passphrase in not a password, rather it is used to encrypt the data. Therefore, it is not possible to change the passphrase for an existing job once it has been set.
Step 2: Backup destination
This is where you can specify where the data will be stored. For scheduled automatic backups, make sure that the backup medium is available at the specified time. Since the URZ object storage is always available, it is suitable as a storage medium. Furthermore, unlike an external (USB) hard drive, it is not integrated into the file system of the computer being backed up and is therefore safe from Trojans.
The following will show you the necessary settings to use the URZ object storage:
- Storage type: Select SFTP (SSH) .
- Server: client-backups.urz.uni-heidelberg.de
- Port: 2222
- Path on server: Please create a new folder for each back up. Creating subfolders is also possible. When backing up multiple computers, we recommend setting the folder path to "computer name/backup name". In our example, pogo15 is the computer name and the backup name is taegliche sicherung. The folder and the subfolder are separated using a forward slash. Only use lowercase letters for the folder names. The folders are created automatically. When you permanently delete a backup job, you can also delete the folder provided that it is otherwise empty.
- Username: Your Uni ID
- Password: The associated password
Please note the following information:
- Port 2222 is only accessible from within the Heidelberg University network. For computers outside the University network, you must establish a VPN connection.
Before clicking Next, please select "Test connection".
Test connection
You can check whether you have configured the settings correctly by clicking Test connection. First, the server will transmit the certificate fingerprint. Please verify this fingerprint. Only click Yes if the following key is displayed:
rsa-sha2-512 4096 0E:49:4D:85:6D:77:26:08:27:3C:FD:E8:5D:C5:C3:E0
If a different fingerprint is displayed, please terminate the process.
If the displayed fingerprint is confirmed by clicking Yes, it will be saved in the configuration file of the backup job and be checked automatically at the start of each future backup.
Step 3: Source files
This is where you will specify the folders or directories to be backed up. You can select the files and folders by checking the box next to the relevant files/folders. Subfolders and the files contained within are included by default.
Only back up the files that cannot be restored by reinstalling/repairing Windows or another software. For instance, let's say you specify the Windows directory C:\Users (or the corresponding macOS or Linux directory) with its files. If Windows stops working properly, there is no point in restoring individual files from the previously backed up directory. This also prevents the backup from becoming needlessly large and placing an unnecessary load on the local computer. Storage space on the object memory is also finite.
If possible, exclude files that do not require a backup: This includes, for example, the local copy of the heiBOX folder (usually called Seafile) since this data is already backed up by the server. This also applies to all other cloud services with a synchronizing function, such as Dropbox, Google Drive, etc. Additionally, .pst and .ost files should always be excluded. Files of these types belong to Outlook, are very large and change constantly, but they do not require a backup if you have an Exchange mailbox or access your emails via IMAP, since all emails are available on the server.
No general statement can be made regarding the backup of the Application Data or Appdata subfolder, as what is stored in this folder depends greatly on how the applications are used.
Files and folders can be excluded from the backup either by deselecting them in the window shown above or by opening the “Filters” menu further down. Here you can exclude files (you can use * as a wildcard, e.g. "*.pst") or specific folders.
It is useful to create different backup jobs with different source data. This way, they can be backed up on different schedules and according to varying levels of importance or frequency of change.
You can select the files by checking the corresponding boxes.
After completing your selection, continue to Step 4 by clicking Next.
Step 4: Schedule
If you would prefer not to start your backup manually, you can set an automatic backup according to a schedule. If the computer is switched off at the specified backup time, the backup will be performed as soon as the computer is back in operation.
Tip: In the settings, you can set a waiting period before a scheduled automatic backup begins. This way, the computer does not have the additional workload of running Duplicati immediately after startup. Duplicati places a heavy load on older computers in particular: Encryption requires a lot of computing power, and the repeated file access places a heavy load on the hard drive. Therefore, it is recommended to select a backup time outside the usual working hours, but please make sure that the computer is turned on at the time of the scheduled backup.
If the computer is located outside the University network, you will need a VPN connection. Especially for a scheduled backup, it is especially important to ensure that this VPN connection is active at the time of the backup.
Complete this step by clicking on Next.
Step 5: General Options
- Remote volume size: The default setting of 50 MB is ideal for the object storage and should not be changed. This 50 MB means that the backup on the object storage is comprised of individual files of 50 MB each.
- Backup retention: We recommend the Smart backup retention for this setting. As the backups get older, they will be thinned out. In the event of an emergency, all existing backups can be used.
Finish creating the new backup job by clicking Save at the bottom of the window. The backup job should now appear on the home screen.
The backup will now be performed automatically according to the schedule--if automatic backups have been set up--and can also be started manually at any time by clicking Run now. You can change the settings at any time by clicking on the job name. You will also find other options, for example the option to delete the job.
Restoring data
You can restore files and folders by clicking Restore on the left side of the home screen. The easiest way to do this is to select a backup job. It will then use the job's database and configuration file. It is also possible to restore the backup job manually if the database and the configuration file are not available. This may be the case if the computer has been restarted after a hardware failure, for example, or if you want to restore the data to another computer. Incidentally, this is also a way to transfer data from one computer to another. However, in this case you will need to enter all required information manually.
After clicking Restore, you will be presented with the following options, as shown in the accompanying image:
Direct restore from backup files: This should be selected if neither the configuration file nor the database are available, e.g. when the computer has been reset or the data is to be restored to another computer. In this case, you must provide all the information about the backup location and the passphrase. The restoration process also takes much longer than when the local database can be used.
Restore from configuration: This option allows you to revert to a previously exported configuration file, and therefore you do not need to provide as much information as in the former option.
All previously created backup jobs will be displayed underneath this option. Restoration from a backup job is the easiest and most common method. We have selected our example "Tägliche Sicherung". Once you have selected your back up job, click on "Next".
Step 1: Select files
In the Restore from field, you can choose which backup generation you want to revert to, if you have chosen Smart backup retention in the backup settings. The most recent backup is set as the default.
In the File search field, you can select which files you want to restore. If a file or folder is missing because you may have deleted (by mistake), you must first select a backup generation in the Restore from field when the file or folder was still present. Only the files/folders that were backed up with the corresponding backup job will be available. Proceed by clicking "Continue".
Step 2: Restoration options
Where should the files be restored to?: Here you can either choose the same location as the backup or specify a different one. If you choose Original location, and the location contains files with the same name, you can select how these files will be processed under How do you want to handle existing files?.
If you do not select Restore Write and Read Permissions under Permissions, the restored files will gain the permissions of the parent directory. This should be done particularly when restoring files that were backed up to another computer. On Linux systems, it may be useful to check this box, because this is the only way to restore execution permissions.
Clicking on Restore will start the restoration process.
The quality of a backup tool only becomes apparent after a successful restoration. Therefore, you should test it before you need to use it!