Comparing IDAS and Argon
- Giang Rudderham
On this page:
The goals of this article are:
To give users a brief overview of the differences and similarities between IDAS and Argon
To help users decide which tool(s) might be suitable for their computational needs
To refer users to additional resources for IDAS and Argon
What are IDAS and Argon?
The Interactive Data Analytics Service (IDAS) and Argon are 2 computing services provided by Information Technology Services - Research Services at the University of Iowa.
IDAS supports large-scale and collaborative data analytics, using interactive tools such as RStudio, JupyterLab, and Jupyter Notebook. IDAS features its own high-performance computing resources and graphics processing units (GPUs). Users access IDAS through a web browser.
Argon is a high-performance computing cluster, where users can submit their computational jobs and request the resources that their jobs need. Users typically access Argon through an ssh (secure shell) client, such as the Terminal in Linux and Mac or SecureCRT in Windows (among other options).
How do I choose between IDAS and Argon for my research project?
IDAS is designed for interactive data analytics, so it is expected that users continuously interact with the programming tools (RStudio, JupyterLab, or the Terminal) while their IDAS session is active. Users can type commands and code in RStudio, JupyterLab, or the Terminal in IDAS, and get a returned output almost instantaneously (unless the code or commands take a long time to execute).
By contrast, since Argon is a batch job system, users typically write a job script and submit their jobs to a scheduler system.
Users may find it easier to get started on IDAS, especially if they are already familiar with RStudio and JupyterLab. By contrast, to get started on Argon, users may need to become familiar with a command-line interface and a job scheduler system.
Some users start by using IDAS for their research projects. Then, as they become more familiar with Argon, or if their computational needs exceed what IDAS can provide, they can switch to using Argon.
The following table provides a quick glance of IDAS and Argon and links to further information.
| IDAS | Argon |
|---|---|---|
Information about the system, including how to request an account, service alerts, etc. | https://research.its.uiowa.edu/our-services/computing-services/argon-high-performance-computing-hpc | |
How to access the system? | Users log in through a web browser - https://notebooks.hpc.uiowa.edu/ | Users typically access Argon through an ssh (secure shell) client, such as the Terminal in Linux and Mac or SecureCRT in Windows, among other options. |
How to run code on the system? | Interactive system. Users type code and commands in JupyterLab, RStudio, or the Terminal and get a returned output. | Batch job system. Users typically write a job script and submit their jobs to a scheduler system. |
What software is available? |
| Many software packages are available. Please see Argon Software List and its sub-pages. |
Quota for a user’s home directory | 100 GB for each research user For more information, please see File Storage in IDAS | 1 TB For more information, please see Home Accounts |
Documentation | Interactive Data Analytics Service Documentation and its sub-pages | Cluster Systems Documentation and its sub-pages |
Use with sensitive data? | Please see the https://its.uiowa.edu/services/protecting-sensitive-data/data-classification-guide-it-services | |
Contact information | For support, contact research-computing@uiowa.edu. You might hear from different team members, depending on whether your questions are about Argon or IDAS. | |
Can I access Argon from IDAS?
IDAS and Argon are 2 separate services with separate computing environments. Users can have both IDAS and Argon accounts. However, those accounts are separate.
Here are some of the things that you can do with Argon while you’re logged in to IDAS:
Mount Argon home directory in IDAS
You can choose to mount your Argon home directory in your IDAS session. This is helpful if you want to access the contents of your Argon home directory while working in IDAS.
Log in to the IDAS at https://notebooks.hpc.uiowa.edu/
In the Server Options form:
If you have an Argon account, select Yes in Mount Argon Home.
Then click Start at the end of the Server Options form.
After your IDAS session has started successfully, the contents of your Argon home directory will be available in the
hpchomefolder in your IDAS account.
From the Terminal in IDAS, the path to the hpchome folder in your IDAS account is /home/HawkID/hpchome, where HawkID is your HawkID.
In your IDAS session, you can move files between your Argon home directory (in the
hpchomefolder) and your IDAS home directory (outsidehpchome).
Additional notes
You can mount your Argon home directory on your local computer, by following the information here: Home Accounts. This allows more flexibility to move files between your personal computer, your Argon account, and your IDAS account.
Prerequisites:
Mount your Argon home directory on your local computer
Mount your Argon home directory in your IDAS session
Possible workflows:
You can move files between your personal computer, your Argon account, and your IDAS account, for example:
Move files from your local computer → your Argon home directory → your IDAS account.
This is helpful if you want to upload larger files to your IDAS account.
Move files from your local computer → your Argon home directory.
This is helpful if you have very large files that do not fit in your IDAS home directory. You can store those files in your Argon home directory, and then access them from the
hpchomefolder in your IDAS account.
Move files from your IDAS account → your Argon home directory → your local computer.
This is helpful if you want to download larger files from your IDAS account to your local computer.
Log in to Argon using the IDAS Terminal
From the IDAS Terminal, you can log in to Argon with the usual ssh command. This is helpful if you want to quickly check on your Argon jobs while logged in to IDAS. However, after logging in to Argon, you will use Argon resources to run your job, not IDAS resources.
Log in to the IDAS at https://notebooks.hpc.uiowa.edu/
Start a Terminal session.
In JupyterLab, click the Terminal tile under Other in your JupyterLab Launcher page to start a Terminal session.
b. In RStudio, click on the Terminal tab, which is next to the "Console" tab in RStudio.
In the Terminal, log in to Argon by running the usual ssh command:
ssh -p 40 HawkID@argon.hpc.uiowa.eduwhere HawkID is your HawkID.
Once you’re logged in to Argon, you can use the usual commands such as
qsub,qstat, etc. to submit and monitor your jobs in Argon. However, at this point, you are using Argon resources to run your job, not IDAS resources.Once you are done with Argon and want to return to using the IDAS Terminal, run:
logoutThis will close your Argon session and return you to your IDAS session. For example:
[grudderham@argon-itf-login-4 ~]$ logout
Connection to argon.hpc.uiowa.edu closed.
grudderham@idas-research-grudderham:~$You can also have more than 1 Terminal session running in IDAS. For example, you can have 1 Terminal session logged in to Argon, and another Terminal session to manage your IDAS home directory.
What is not shared between IDAS and Argon?
Here are some of the things that Argon and IDAS do not share:
Argon and IDAS have separate software stacks. So, for example, Argon and IDAS will have different software and different versions available.
Your Argon home directory is separate from your IDAS home directory. However, you can mount your Argon home directory in your IDAS session as described above.
Argon and IDAS have separate hardware (CPUs, GPUs, etc).
Contact
If you have any questions or comments, please contact research-computing@uiowa.edu.