Quest Analytics Nodes#
The Quest Analytics Nodes provide RStudio, Jupyter, and SAS Studio interfaces directly in an internet browser. They use the same data storage system as Quest. They are suitable for workflows that require low or moderate computational resources.
Important
Due to the finite compute resources dedicated to the Quest Analytics Nodes, the amount of compute resources available to each user is limited. If your session takes up too many resources, it will be automatically stopped by our system. For the larger interactive jobs, Quest OnDemand is a good alternative.
Connect#
To use the Analytics Nodes, users must:
Have an active Quest allocation
Be on the Northwestern network; off-campus users must use the GlobalProtect VPN to connect.
Users can connect using any web browser. Go to:
Jupyterhub: https://jupyterhub.quest.northwestern.edu
SAS Studio: https://sasstudio.quest.northwestern.edu
Log in with your NetID and password.
Usage Guidelines#
The Quest Analytics Nodes are suitable for small and moderately sized interactive jobs only. Users running computationally intensive analyses may have their sessions cancelled or be asked to convert their work to run as a scheduled job on Quest. Users who generate exceptionally large temporary files may be contacted to alter their workflow. Sessions generally time out after one hour of inactivity.
SAS Studio
Limited to 4 GB of RAM (memory)
Jupyter
Limited to 8 cores
Please limit memory use for large jobs
RStudio
Limited to 8 cores
Please limit memory use for large jobs
Users utilizing a large amount of memory, especially those using over 120GB of RAM, may be asked to move their analysis to other systems. Researchers requiring high memory limits for analysis should contact quest-help@northwestern.edu to discuss options.
Important
Users of the Quest Analytics Nodes are subject to the same guidelines and restrictions as all other users of Quest. Note in particular that Quest is not approved for storage or analysis of restricted data, and it may not be used for commercial purposes.
Note
Users can submit jobs to Quest through the Slurm scheduling system from the Quest Analytics Nodes.
Software#
The JupyterHub version is 5.2.1 and the default environment contains a limited number of packages: numpy, pandas, matplotlib, scipy, scikit-learn, scikit-image, and scipy. Users can create their own conda environments with different Python versions or libraries installed (see Accessing Conda Environments in Jupyter for more details).
The RStudio Server version is 2024.12 and runs with R version 4.4.0. The Quest Analytics Nodes run the free, Open Source Edition of RStudio server, which does not allow users to individually select their version of R from within RStudio or run multiple R sessions simultaneously.
SAS is version 9.4 with maintenance release 4 applied (STAT module is version 14.2). Packages are restricted to those available in the Education Analytical Suite .
Hardware#
Quest Analytics Nodes
4 nodes total
Each node: 64 CPU cores and 2 TB of RAM
-
All your files, packages, and settings are available from all nodes
Cannot directly connect to a specific node
If you refresh the page or lose connection, you may be reconnected to a different node and lose access to your previous session
GPUs are not available as part of Quest Analytics
The Quest Analytics Usage Tips page provides additional detail about specific use cases on the Analytics Nodes.