Data Analysis Gateway or DAG provides a set of interactive data analysis nodes for intermediate computation, which can be completed in a short timeframe.
This means that any spawned instance is limited to 2 hours of inactivity before it will be terminated. DAG instances have access to 8 compute cores and 8GB of memory

The spawned instances are non-persistent, meaning that any change made during a session is lost once the server is terminated.
The only exception to this is the data that is saved in the provided mount directory (i.e. ~/work)

For more information about how you can ease the task of configuring your instances and our future roadmap for allowing customization, check out the FAQ section "How do I install and run software XYZ in Jupyter?" at ERDA FAQ