Versions Compared

Old Version 15

changes.mady.by.user Cody Johnson

Saved on

compared with

New Version Current

changes.mady.by.user Giang Rudderham

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Info

To request an IDAS account for research use, please fill out the following Workflow form: https://workflow.uiowa.edu/form/idas-account. 

Jupyter Notebook for Python, R, and Julia

Once you get access as an IDAS user, here is the process for accessing the system:

Please also see our policies for IDAS account requests in Requesting An IDAS Account.

On this page:

Table of Contents

...

Logging in to the Research instance on IDAS

Once you receive an email confirming that your IDAS account is ready, follow the steps below to start an IDAS session.

1. Open any web browser. 

2. Go to https://

...

idas.uiowa.edu/research/hub/home. 

...

3. Log in using your HawkID and password

...

  1. image2019-12-18_16-41-3.pngImage Removed
  2. Selecting HPC home will mount your Argon home and allow you to access the files stored there.
  3. Any LSS shares entered into the LSS fields will also be mounted for use by the IDAS instance.
    1. If you instance fails to spawn please check that the share names are correct.
  4. Please note that Python, R, and Julia will each have its own container image. For example, to start a Julia instance, please select one of the three options for Julia under IDAS Image in the Spawner form.
  5. Image Removed

...

.

4. Authenticate the login with Duo. More information about Duo can be found here.

5. Click the blue button Start My Server.

...

Troubleshooting notes

Instead of Start My Server, do you see a red button Stop My Server, like in the screenshot below?

image-20241017-184242.pngImage Added

This means that your IDAS session is already running. From here, you have 2 options:

6. Fill out the Server Options form, using the instructions on the form. Please also see the notes below.

a) CPU Cores: Fill in the number of CPU cores you want to use in IDAS, with 2 as the minimum and 32 as the maximum.

  • If left blank, the default 2 will be used.

b) Memory Limit: Fill in the amount of RAM (GB) you want to use in IDAS, with 8 as the minimum and 512 as the maximum.

  • If left blank, the default 8 will be used.

c) GPU: 

  • If you are not using GPU, choose "None". This is the default selection.

  • If you have GPU-enabled code, select the type of GPU that you want to use. For more information, please see Using GPUs on IDAS.

  • GPUs are a finite resource on IDAS and assigned on a first come first served basis. Please request a GPU only if you need to use it and shut down your IDAS server when you are finished so that other users can share the resources.

d) Mount Argon Home: 

e) LSS Shared:

  • If you have Large Scale Storage (LSS) shares that you want to use with your IDAS session, enter their paths under "LSS Shared" or "LSS Dedicated". Leave these fields blank otherwise. 

    • Please check that the share names are correct.

    • Multiple shares can be separated by a comma. For example: SHL-BUG,SHL-NBS

f) LSS Dedicated:

  • If you have Large Scale Storage (LSS) shares that you want to use with your IDAS session, enter their paths under "LSS Shared" or "LSS Dedicated". Leave these fields blank otherwise. 

    • Please check that the share names are correct.

    • Multiple shares can be separated by a comma.

g) IDAS Image:

  • Under IDAS Image, choose the programming language that you want to use with IDAS. IDAS currently has multiple versions of R, Python, and Julia available

  • If you are not running code that needs CUDA, you do not need to choose a CUDA version.

Under IDAS Image, what’s the difference between the R and RStudio instance types?

  • Choose an RStudio option, such as IDAS-R-RStudio-4.4.1, to start RStudio.

  • Choose an R option, such as IDAS-R-4.4.1, to start JupyterLab. After the IDAS session has started:

    • Click the R tile under Notebook to start a Jupyter Notebook with R.

    • Click the RStudio tile under Notebook to start RStudio.

For more information, please see RStudio in IDAS.

image-20240624-211722.pngImage Added

7. Click the Start button. A new IDAS session will be created. This may take several seconds.

8. After the session has been created, you will see your IDAS home directory. If this is your first time accessing IDAS and you haven't uploaded any files, your home directory will be empty.

9. For more information about using IDAS, please see Using IDAS. 

10. Once you finish your job with the IDAS instance, it is recommended that you manually stop the serversession, please shut down the session, so that you can quickly release the resources you have been using and get them ready for use by other users. You can stop the instance by clicking the Control Panel button at the top right corner and then clicking the Stop My Server button. You will not lose any files by doing so, as long as you save them.

Your session will automatically end after being idle for 18 hours. The session will not end if being active. 

RStudio

To start an RStudio instance, please select one of the three options for R under IDAS Image in the Spawner form. RStudio is available with the GPU instances as well.

Image Removed
Doing so will launch an R instance. In order to launch RStudio, click New, then select the RStudio option.
Image Removed
This will launch an RStudio session, which looks very similar to the local version of RStudio.
Image Removedresources could be used by other IDAS users. For more information about stopping or restarting your IDAS session, please see https://uiowa.atlassian.net/wiki/spaces/hpcdocs/pages/76514933/Accessing+IDAS+for+Research+Use#Restarting-your-IDAS-session.

Warning

Your IDAS session will be automatically terminated if it has been idle for 18 hours with no activities.

IDAS was originally designed for interactive work, so “activities” include logging into IDAS, clicking on the IDAS tab, and interacting with the IDAS interface in some way, such as uploading, downloading, and managing your files. Having code running doesn’t count as an interactive activity. If a user hasn’t logged into IDAS for 18 hours, their session will be automatically stopped to release resources for other users.

Restarting your IDAS session

When to restart your IDAS session?

You might want to restart your IDAS session in one of the following scenarios:

  • You want to change the CPU cores, RAM, or GPU available for your IDAS session.

  • Your current IDAS session isn’t connected to your Argon home directory or LSS shares, and you want to mount your Argon home directory or LSS shares in your IDAS session.

  • You want to change to a different IDAS image, such as:

    • changing from Python to R or Julia

    • changing to a specific software version, such as changing from R 4.0 to R 4.4

    • changing to an IDAS image that has CUDA, such as changing from R 4.4 without CUDA to R 4.4 with CUDA.

  • You are working with our team to troubleshoot an issue, and we have recommended that you restart your IDAS session.

  • You are finished with your tasks on IDAS. In this case, shutting down your IDAS session will release resources to other users. This is especially helpful if your session was using a GPU, since GPUs on IDAS are limited and available on a first come first served basis.

Steps to restart your IDAS session

1. Save any changes to the scripts, datasets, or files that you want to save. 

2. In your web browser, navigate to https://idas.uiowa.edu/research/hub/home

3. Click the red button Stop My Server.

Screenshot 2024-01-16 133249-20240116-183249.pngImage Added

4. Wait a few seconds for the blue button to change to Start My Server.

Screenshot 2024-01-16 133207-20240116-183207.pngImage Added

5. Click Start My Server to log in to IDAS again if you want to continue using IDAS. Otherwise, if you are done with IDAS, close your browser tab. 

Switching between JupyterLab, Jupyter Notebook, and RStudio

The following interfaces are available in IDAS:

  • JupyterLab - available in all IDAS instances

  • Jupyter Notebook - available in all IDAS instances

  • RStudio - only available in the R and RStudio instances

IDAS users can quickly switch between JupyterLab, Jupyter Notebook, and RStudio by changing the URL in their web browser.

Research users can use the following URLs.

  • To access JupyterLab - Please fill in your HawkID:

Code Block
https://idas.uiowa.edu/research/user/HawkID/lab

  • To access Jupyter Notebook - Please fill in your HawkID:

Code Block
https://idas.uiowa.edu/research/user/HawkID/tree

  • To access RStudio (only available in the R and RStudio instances) - Please fill in your HawkID:

Code Block
https://idas.uiowa.edu/research/user/HawkID/rstudio/

Info

JupyterLab, Jupyter Notebook, and RStudio are different interfaces you can use to access your IDAS account. You still have 1 IDAS account, and all your files and folders are still the same no matter which interface you use.

The following table shows the instance types in IDAS, their corresponding default interfaces, and the interfaces that are available.

Instance Type - When you log in to IDAS, you can choose an IDAS Image in the Server Options form. There are 4 main instance types. Each of them has multiple language versions and CUDA versions available.

Default Interface - This is the default interface that you will see after your IDAS session has started successfully.

Available Interfaces - These are the interfaces that are available for this instance type. You can switch between them in your IDAS session. You will still have access to your IDAS account with the same files; only the interface will be different.

JupyterLab available?

Jupyter Notebook available?

RStudio available?

Python

JupyterLab

yes

yes

no

R

JupyterLab

yes

yes

yes

RStudio

RStudio

yes

yes

yes

Julia

JupyterLab

yes

yes

no

Contact

If you have any questions or comments, please contact research-computing@uiowa.edu.