What Do XNAT Admins Need to Know?
XNAT is now easier to set up and configure than ever before, with more control being brought in front of admins via the Admin UI.
With that said, however, there are some tasks related to administering your XNAT that still require access to the file system (such as adding pipelines and plugins), and still some reporting (such as complex data freezes) that can only be done with access to the PostgreSQL database.
These guidelines will teach you the "what", "why", and "how" of being an XNAT Admin.
To get a guided walkthrough of the core tasks of administering an XNAT, try taking the free Admin 101 Course on XNAT Academy
XNAT First-time Setup
The first time you launch XNAT, there are a series of initial configurations that must be set. This will only happen once.
Getting To Know XNAT
XNAT is a powerful set of software with a complex set of components under the hood. However, we provide administrative tools that allow you to control many core behaviors directly from the UI. You must be logged in as a site administrator to access these controls.
Managing Users and Roles
Managing Sites and Studies
You can set XNAT Administrator privileges to give admin users access to all studies and study data in your XNAT. For more details, see Administering Users.
Adding Data Types and New Functions to XNAT
Extending Data Types with Custom Forms
Administering Image Data Import Methods
- Admin UI - Session Upload and Anonymization Settings
- Configuring the Use of the AutoRun Pipeline
- Connecting XNAT to DICOM Scanners and PACS
- Custom DICOM Object Identifiers
- DICOM SCP Receiver Configuration
- Enabling Snapshot Generation on Compressed Images
- How To Use a Project ID as a DICOM AE Title
- XNAT Custom DICOM Routing Rules in the Admin UI
Managing Resource Files
Computing in XNAT
The Pipeline Engine is being deprecated in XNAT 1.8, and the "Autorun" pipeline is no longer running by default. Snapshot generation is now handled by an internal service. Processes can still be run in pipelines, but in the interest of reproducible science, we highly recommend converting your computational processes to containers.