CyVerse_logo

Home_Icon Learning Center Home

VICE Documentation

The Visual and Interactive Computing Environment (VICE) is the latest feature in CyVerse’s Discovery Environment (DE) for running interactive applications.

User Guide instructs users on basic functions of the DE: (a) how to launch an interactive app, (b) bring your data into the container, (c) run an analysis, and (d) save or upload your results to the cloud.

Developer Guide gives advanced tips about how to (a) build your own VICE-ready containers, (b) host them on a public registry, and (c) integrate them into VICE and make them available to your user community.

Last, we provide a brief list of featured VICE apps in the DE.

CyVerse_logo

Home_Icon Learning Center Home

About

VICE stands for Visual Interactive Computing Environment and is a part of CyVerse Discovery Environment (DE).

VICE allows verified CyVerse users to launch customized containers of their favorite data science platforms.

What is the big idea?

The DE supports executable applications which run as workflows on high performance or high throughput computing environments. VICE introduces Graphic User Interfaces (GUI) and common Integrated Development Environments (IDE) such as Project Jupyter, RStudio, & Remote Desktops to the DE.

Note

How is VICE different from other DE apps?

Applications or “Apps” in the DE are classified as:

  • executable - meaning the user selects parameters and data for a particular analysis, and submits the job for execution (HTCondor , Tapis, OSG).

  • HPC - executable apps which run on Texas Advanced Computing Center Tapis as high performance computing (HPC) jobs.

  • OSG - executable apps which run on the OpenScienceGrid as high throughput (HTC) jobs.

  • VICE - interactive, visual computing environments with GUI or IDE.

CyVerse_logo

Home_Icon Learning Center Home

# Verified User Information

To access the interactive apps in the Disocvery Environment, we ask users to provide sufficient information about themselves in their user profiles to ensure proper use of the platform.

We ask users to provide a valid email address from an organization, i.e., .org, an educational institution with the .edu ending, or a government .gov – we will not grant access to commercial email addresses, e.g., @gmail.com @yahoo.com @msn.com etc.

We also ask that users provide their |ORCID| – ORCID is the most widely adopted digital identifier for research in the world.

Prerequisites

Prerequisite

Preparation/Notes

Link/Download

CyVerse account

You will need a CyVerse account to complete this exercise

Register

Discovery Environment access

You will get access to Discovery Environment by default. Check the ‘My Services’ tab; if Discovery Environment is not listed, click the ‘Available’ tab, locate Discovery Environment and click the link to request access.

CyVerse User Portal

Fix or improve this documentation:

CyVerse_logo

Home_Icon Learning Center Home

Applications

Currently, VICE apps are categorized as:

  1. Integrated Development Environments (IDE) (e.g., rstudio-verse-latest jupyter-datascience-latest)

  2. Interactive Apps (e.g., Shiny, WebGL, HTML5)

  3. Virtual Desktop Environment (e.g., Ubuntu Desktops w/ Apache Guacamole, VNC, Xpra)

1. What is JupyterLab?

JupyterLab is an interactive development environment for working with notebooks, code and data.

JupyterLab enables you to use text editors, terminals, file viewers, and other custom components side-by-side with notebooks in a tabbed work area. JupyterLab provides a high level of integration between notebooks, documents, and activities:

  • Drag-and-drop to reorder notebook cells and copy them between notebooks

  • Run code blocks interactively from text files (.py, .R, .md, .tex, etc.)

  • Link a code console to a notebook kernel to explore code interactively without cluttering up the notebook with temporary scratch work

  • Edit popular file formats with live preview, such as Markdown, JSON, CSV, Vega, VegaLite, and more

1.1 What is a Jupyter Notebook?

The Notebook (formerly IPython Notebook) is Project Jupyter’s flagship project for creating reproducible computational narratives.

It enables users to create and share documents that combine live code with narrative text, mathematical equations, visualizations, interactive controls, and other rich output.

Notebook documents (or “notebooks”) are documents produced by the Jupyter Notebook App, which contains both computer code (e.g., python, r, julia) and rich text elements (paragraph, equations, figures, comments, images, links, etc.).

1.2 JupyterLab VICE

The JupyterLab VICE apps are based on official Project Jupyter Images integrated into the CyVerse Discovery Environemnt (DE).

2. What is RStudio?

RStudio is a free and open source integrated development environment for R, a programming language for statistical computing and graphics. Some of its features include:

  • Customizable workbench with all of the tools required to work with R in one place (console, source, plots, workspace, help, history, etc.)

  • Syntax highlighting editor with code completion

  • Execute code directly from the source editor (line, selection, or file)

  • Full support for authoring Sweave and TeX documents

  • Runs on all major platforms (Windows, Mac, and Linux) and can also be run as a server, enabling multiple users to access the RStudio IDE using a web browser

2.1 RStudio VICE

The RStudio VICE apps are based on official Rocker Project Images.

3. What is Shiny?

Shiny is an open source R package that provides an elegant and powerful web framework for building web applications using R. Shiny helps you turn your analyses into interactive web applications without requiring HTML, CSS, or JavaScript knowledge. Some of its features include:

  • Build useful web applications with only a few lines of code—no JavaScript required

  • Shiny applications are automatically “live” in the same way that spreadsheets are live. Outputs change instantly as users modify inputs, without requiring you to reload your browser

  • Shiny user interfaces can be built entirely using R, or can be written directly in HTML, CSS, and JavaScript for more flexibility

  • Works in any R environment (Console R, Rgui for Windows or Mac, ESS, StatET, RStudio, etc.)

  • Attractive default UI theme based on Twitter Bootstrap

  • A highly customizable slider widget with built-in support for animation

  • Pre-built output widgets for displaying plots, tables, and printed output of R objects

  • Fast bidirectional communication between the web browser and R using the websockets package

  • Uses a reactive programming model that eliminates messy event handling code, so you can focus on the code that really matters

  • Develop and redistribute your own Shiny widgets that other developers can easily drop into their own applications

3.1 Shiny VICE

The Shiny VICE apps are based on official Rocker Project Images.

Shiny Apps require the contents of a Shiny App be added as an input folder prior to launching a VICE app - otherwise the default Shiny App demo will be the only contents in the container.

Customized Shiny Apps can be created and deployed using the Tool and App builder.

4. What is Ubuntu Desktop?

The Ubuntu Desktop has a full Guacamole installation and Ubuntu XFCE desktop. This allows users to have a simple all-in-one desktop through their web browser. Users can run any interactive or visualization tool that can run on the most recent linux distros. Solutions to support the inevitable array of linux applications that user will want. Potential options include:

  • Separate image per application

  • Network fs (e.g. NFS, Ceph, etc) containing all applications

  • Per-application network fs

  • On-demand installation of application via script/ansible

4.1 Ubuntu Desktop VICE

Linux Desktops using Apache Guacamole, and Xpra have been integrated into the DE.

Fix or improve this documentation:

Home_Icon Learning Center Home

Cloud Shell

1. Log into Discovery Environment

Log into CyVerse DE with your CyVerse credentials.

If you have not yet created an account, go to the User Portal and do so.

The Cloud Shell App is already pinned to the Discovery Environment Tool Bar (left side of screen):

Alternately, use the search to specify ‘Apps’ then type keyword Shell or Cloud Shell

You can see Cloud Shell listed when you expand the three bars Menu icon.

cloudshell-0

2. Launch analysis

cloudshell-1

Instant Launches start the App immediately without the opportunity to add any input data or increase the VM size.

Clicking on the Apps will show the Cloud Shell in “Featured Apps”.

By default the Analysis Name is the name of the App plus the date and time.

The output folder will use this same name and be written to the /iplant/home/username/analyses/ directory AFTER the interactive analysis is completed.

Note

Deleting a running Analysis will stop it from writing outputs to your data store analyses/ directory.

Note

The input files and/or folders can be selected under the ‘Parameters’ tab.

Tip

If you have an existing set of files or directories, you can import these into the app using input files and/or folder in the Parameters .

4. Activate a conda environment

The Cloud Shell comes with multiple languages and package managers pre-installed. These include go, python, and rust.

Package managers include conda and cargo in addition to linux apt and apt-get.

Your identity inside the container is user and you have limited sudo privileges to install new packages into the container.

These changes are not saved after the analyses ends, or when you start a new Cloud Shell Analyses later.

Tip

The cloud shell is running a terminal multiplexer called tmux which keeps your session active even after you’ve closed your browser tab.

tmux keyboard commands should function normally.

To activate conda:

` conda init `

and then

` conda activate base `

If you recieve a message about refreshing your screen, you can exit the cloud shell by typing “exit” and then hitting the refresh button on your browser tab.

5. Adding data to your analysis

To connect to the CyVerse DataStore, you can initiate an iRODS iCommands iinit

You should now be connected to your /iplant/home/username home directory.

ils

or

` ils /iplant/home/username/ `

To view the ‘shared’ directory try:

` ils /iplant/home/shared `

Download data into your Cloud Shell with iCommands by running iget

` iget -KPbvrf /iplant/home/shared/cyverse_training/ `

```

6. Complete and Save Outputs

After finishing your analyses, you can save the outputs back to Data Store.

You can either use iput to copy your new files back to your user space, or if you’ve left your new work in the /home/user/work folder, it will be copied back to your /iplant/home/username/Analyses/ directory.

You can find the outputs you generated (if any) using the same steps as before, but this time select the ‘Go To Output Folder’.

7. Terminate your app

The Discovery Environment is a shared system. In fairness to the community, we ask uses to “Terminate” any apps they have started when they are no longer running analyses.

In the Analyses window, select the app with the checkbox, then select “More Actions” and “Terminate” to shut the app down.

Any new data in the /home/user/work directory will begin copying back to your folder at this time.

Any input data which you added when the app started using the conventional launch feature will not be copied.

Warning

VICE apps only run for a pre-determined amount of time, typically between 4 and 48 hours. If you have opted for email notifications from DE, then you’ll get a notification 1 day before and 1 hour before the app gets terminated. If you want to extend the time, you need to login to https://de.cyverse.org, find your analysis and then click the hour glass which automatically extends the app run time.

Fix or improve this documentation

Home_Icon Learning Center Home

JupyterLab

1. Search for JupyterLab

Log into CyVerse DE

Use the search bar to specify ‘Apps’ or type keywords Jupyter or JupyterLab Datascience

jupyter2-0

2. Launch analysis

jupyter2-1

Instant Launches will start the App immediately.

Clicking on Apps find the JupyterLab you’re interested in and then click on the blue run button.

By default the Analysis Name is the name of the App, and the Output Folder is where any work done in the container will be written to /iplant/home/username/analyses/analysis_name/ directory AFTER the interactive analysis is completed.

Note

Deleting a running Analysis will stop it from writing outputs to your data store analyses/ directory.

jupyter2-2

Note

The input files and/or folders can be selected under the ‘Parameters’ tab.

Tip

If you have an existing Jupyter workbook, you can import it into the app using input files and/or folder in the Parameters .

4. Create Jupyter notebook

jupyter2-3

Jupyter notebooks (.ipynb) combine code with narrative text (Markdown), equations (LaTeX), images and interactive visualizations.

To create a notebook, click the + button, this opens the new Launcher tab.

The JupyterLab Datascience containers have three pre-installed kernels: Python3, Julia, and R.

Official Jupyter Notebooks

Tip

To open the classic Notebook view from JupyterLab, select “Launch Classic Notebook” from the Help Menu.

5. Adding data to your analysis

To connect to the CyVerse DataStore, click the little CyVerse orb in the left side of the Lab.

jupyter2-4

You should now be connected to your /iplant/home/username home directory. Navigate to the ‘shared’ directory by clicking one order higher on the /home directory, you should now see your username and the /shared path.

Navigate to /shared/cyverse_training/platform_guides/discovery_environment/jupyterlab and open the Penguins sample dataset

jupyter2-5

Note

There are plenty other cool stuff that you can do in Jupyter Lab such as using consoles, using terminal and using text editor.

6. Write your own code

Once you open a new notebook, you can start writing your code, put markdown text, generate plots, save plots etc.

jupyter2-6

6. Complete and Save Outputs

jupyter2-7

After finishing your analysis, you can save outputs to Data Store by clicking the Analysis window, then select the VICE analysis that you are running and select Terminate under the “Analyses” button.

jupyter2-8

After you had done this, you can find the outputs that you generated (if any) using the same steps as before, but this time selecting ‘Go To Output Folder’.

Warning

VICE apps only run for a pre-determined amount of time, typically between 4 and 48 hours. If you have opted for email notifications from DE, then you’ll get a notification 1 day before and 1 hour before the app gets terminated. If you want to extend the time, you need to login to https://de.cyverse.org, find your analysis and then click the hour glass which automatically extends the app run time.

Fix or improve this documentation

Home_Icon Learning Center Home

RStudio

RStudio is a free, open source IDE (integrated development environment) for R. Its interface is organized so that the user can clearly view graphs, data table, R code and ouput at the same time. It also offers an Import-Wizard-like feature that allows users to import CSV, Excel, SAS (.sas7bdat), SPSS (.sav), and Stata (*.dta) files into R without having to write the code to do so.

1. Search for Rstudio App

First log-on CyVerse DE

Use the search bar to specify ‘Apps’ or search for ‘rstudio’.

rstudio3-1

2. Launch analysis

Find the RStudio app you are interested in and click on the app. Change the analysis name and output folder or leave it to defaults.

rstudio3-2

Launch the Rstudio app by selecting an example folder or click on the details to quick launch example data. You can select different input files and/or folder.

Tip

You can use input files to import a script into the app.

rstudio3-3

Use quick launch to input example data… rstudio3-4

Note

You will not see any files when selecting the folder if you selected input by folder. Rest assured that they will be there once the app begins to run.

Launch the analysis after you are finished selecting the input files (if any). rstudio3-5

4. Launch Rstudio

After you start the VICE App, you will be taken to a loading screen.

rstudio3-7

5. Write/Run your code

In the Rstudio script section, you can write your code, generate plots, save plots etc.

Tip

As a first step, check that the files you wanted to import are in the app. Go to the bottom right of the app, and check under ‘Files’ for your files.

rstudio3-8

6. Complete and Save Outputs

Complete your analysis by clicking the Analysis window, then select the rstudio analysis. From More Actions tab, click ‘Terminate’.

rstudio3-9

rstudio3-10

After you had done this, you can find the outputs that you generated (if any) using the same steps as before, but this time selecting ‘Go To Output Folder’.

Warning

Currently, VICE can run for 48 hrs beyond which the apps will be terminated. So make sure you run your analysis before 48 hrs.

Fix or improve this documentation

Home_Icon Learning Center Home

Sharing VICE apps with collaborators

You can share your running VICE workspace with colleagues (with a CyVerse account) who can see and edit your notebooks, logs, and outputs.

To share your running workspace

Click on the Analyses

vice-share-1

Select the running analysis and click on Share from the top bar.

vice-share-2

From the sharing window, search your collaborators by CyVerse username, email or group name.

vice-share-3

Opening workspaces shared with you

Click on the Analyses

vice-share-1

Select the running analysis and click on Go to analysis (square arrow box on right).

vice-share-6

This will launch the shared analysis in a new window.

vice-share-7

There are also different ways to share a VICE app without sharing a running instance of that app.

  • Apps that have been made public in the Discovery Environment can be shared with the public app’s URL.

  • Unpublished apps (those in your private workspace that have not yet been made public) can be shared with specific users or teams. See Sharing your App or Workflow and Editing the User Manual.

  • A Quick Launch configuration can be created, and then the URL to the Quick Launch can be shared.

Creating and sharing Quick Launch configurations

Quick Launches provide a way to set default parameter values for an analysis, which can make it much easier to launch similar jobs without having to select the parameter values that the jobs have in common for every new analysis.

To create a Quick Launch, open the app as if you are going to run it, fill in any default parameters for this Quick Launch, then click the Create Quick Launch button instead of the Launch Analysis button.

quick-launch-create

Then you can name the Quick Launch and make it public.

quick-launch-dialog

To share the link for a public Quick Launch, or to copy its badge code for embedding in web pages or in repo markdown files, click the Quick Launch menu item from the app’s 3-dot menu in the Apps window.

quick-launch-use

Note that private Quick Launch configurations cannot be shared, and clicking on them from this listing simply opens the app launch dialog with its parameters pre-filled, according to the Quick Launch configuration.

If you don’t need to save any default inputs or parameters, you can also use the app’s public URL for sharing the VICE app, which acts the same as a Quick Launch, but it doesn’t have any parameters pre-filled.

Fix or improve this documentation:

CyVerse_logo

Home_Icon Learning Center Home

Workflow

CyVerse hosts many popular data science applications, e.g. Jupyter Lab, RStudio, and Shiny. These applications can be started in the Discovery Environment, and the researcher can install additional packages to the running application.

In cases where the installation may be complex, long, or require additional system administrator level access, the researcher can use the existing CyVerse container as a base image for their own new container. The researcher can add their own packages and then deploy the new app in the Discovery Environment.

flow

Fix or improve this documentation

CyVerse_logo

Home_Icon Learning Center Home

Guidelines for adding interactive tools in DE

Prerequisites

Adding VICE Tools in DE is different from non-interactive Tools. VICE applications like Jupyter and RStudio run on an open port for enabling their web UI.

  1. Ensure that the listen port for the web UI has a sane default and is set in the Dockerfile.

  2. The working directory is set

  3. All commonly needed dependencies are installed in the container image - you will not have root privileges later

  4. The default user set

  5. Disable any additional authentication (CyVerse provides CAS authentication and authorization).

  6. URLs will work sanely behind a reverse proxy. If they don’t, you may need to add nginx to the container.=

Community images as your base image

If you need to set the configurations at all (see above), you’ll need to create a new Dockerfile that uses the community-provided image as a base. Your new Dockerfile should deal with custom configurations and dependency installations.

See some examples of VICE apps that uses community images as base image in the Dockerfile

Test your Docker image

Since the images are built based using Dockerfile, make sure you test the Dockerfile before providing it to us. Dockerfile must have Entrypoint. If you cannot provide us the Dockerfile, you can request integration of the app by doing a tool request.

Fix or improve this documentation:

Home_Icon Learning Center Home

Building DE tools and apps

Once you build your Docker image (following the guidelines), the next step is building the Tool.

For this you’ll need a Docker image name, any port numbers PORT, User ID UID, working directory WORKDIR, and ENTRYPOINT.

Docker images

The Docker image of your tool is mandatory and it should be available on public registries such as Dockerhub or quay.io. Alternatively you can provide us the Dockerfile and we will build the Docker image for you. If there is no Dockerfile for the tool that you are interested in, then tell us what tool you are interesting in making us as VICE app.

Add Tool in DE

The final step in building the VICE tool is to fill up the “Add Tool” form in DE.

Brifely here are the following steps.

  • Log in CyVerse Discovery Environment

  • Click on the Apps window and click Manage Tools button on the far right hand side of the window

  • Click on Tools button and then finally Add Tools button

You’ll see a Add Tool form like this

|add-tool|

  • Tool name is the name of the tool. This will appear in the DE’s tool listing dialog. This is mandatory field. Eg. “jupyterlab-circos”

  • description is a brief description of the tool. This will appear in the DE’s tool listing dialog. Eg. “Circos is a software package for visualizing data and information that was created by Martin Krzywinski”

  • version is the version of the tool. This will appear in the DE’s tool listing dialog. This is mandatory field. Eg. “1.0”

  • Image name is the name of the image specifier minus the image tag. The image must exist on Dockerhub or quay.io. This is mandatory field. E.g “fomightez/circos-vice”

  • Tag is the image tag. If you don’t specify the tag, the DE will look for the “latest” tag which is the default tag.

  • Docker Hub URL is the url of the image on the Dockerhub. E.g https://hub.docker.com/r/fomightez/circos-vice

  • Type is the type of Tool. For VICE apps, chose “interactive”.

  • OSG Image Path is path of the image on the OSG. You can skip this for interactive tools.

  • Entrypoint is the Entrypoint for your tool. Entrypoint should be present in the Docker image and if not, you should specify it here.

  • Working Directory this is the working directory of the tool and must be filled in with the value you gathered above. E.g /home/jovyan/vice

  • UID is a number and must be filled in with the value you gathered from above. E.g 1000

  • Max CPU Cores is the number of cores for your tool. Eg. 16

  • Memory Limit is the memory for your tool. Eg. 64 GB

  • Min Disk Space is the minimum disk space. Eg. 200 GB

  • Container Ports must be a list of maps with only a single entry. The key in that entry must be container_port and should be filled in with the number value you gathered above.

Warning

It is strongly recommended you do not set the bind to host as true for your added ports when creating a new App**

Creating VICE app for your tool

To create a new app, follow the instructions in here

Important

For VICE apps, be sure to check the box “Do not pass this argument to the command line” for each option you add (for VICE, this is usually just input files and folders.

|vice-do-not-pass|

Fix or improve this documentation:

CyVerse_logo

Home_Icon Learning Center Home

Instant Launches

From the Home tab there are several apps that have an Instant Launch feature which allows you to start the app in a single click.

These apps launch with their default number of cores, amount of RAM, and timeout, and without input data. You can always import data using https protocols or iCommands after launch.

Quick Launches

Quick launch buttons are directed URLs that allow you to share an app with pre-set configurations. After selecting an app, you will be taken to the app launcher where you can select input data sets, and set size and time parameters. Any public app can have a quick launch URL generated for it.

To use one of the Featured Launches listed below in the table, copy the badge (button link) to add to where ever you collaborate (a webpage, project notes, documentation, etc.).

To create your on Saved Launch, proceed with launching the app you want to use. This will be a good time to Favorite the app. In the “Review & Launch” panel, click the “Create Saved Launch” button. You will be asked to name your Saved Launch and check the box when prompted if you would like it to be public. Remember which app you saved, You will find the link under Details of the app you saved.

App Name

Dockerfile

Saved Launch

Rocker Project RStudio Verse

GitHub

rstudio-verse-latest

Rocker Project RStudio Geospatial

GitHub

rstudio-geospatial-latest

JupyterLab Datascience

GitHub

jupyter-datascience-latest

JupyterLab Geospatial

GitHub

jupyter-geospatial-latest

You can design your own badge at Shields.io.

Fix or improve this documentation

CyVerse_logo

Home_Icon Learning Center Home

Frequently Asked Questions

  1. What happens if my VICE app has been running for more than 48 hours?

  2. Can I extend the 48hr time limit on the VICE app?

  3. Can I request the CyVerse team to build the VICE app for my interactive tool if I don’t have Docker image?

  4. I’m getting this error (or similar) with my docker file:

You must set a unique PASSWORD (not ‘rstudio’) first! e.g. run with: docker run -e PASSWORD=<YOUR_PASS> -p 8787:8787 rocker/rstudio

A: Make sure the environmental password is set for rStudio: ENV PASSWORD “rstudio1”. You can also try using this base image for rStudio: cyversevice/rstudio-base:latest or cyversevice/rstudio-verse:3.6.0. For bioconductor images, use upendradevisetty/bioconductor:1.0.

Fix or improve this documentation:

CyVerse_logo

Home_Icon Learning Center Home

VICE Best practices

  1. Smaller docker containers are better. Larger images take longer to transfer and load.

  2. Use our base images for complex programs that require additional configuration files other than the Dockerfile.

  3. Create robust documentation with as much metadata as possible.

Fix or improve this documentation:

Tool Troubleshooting

1. Get the port

You’ll need to figure out the port that the tool uses to present its web interface. This is mandatory and you can integrate a tool without knowing the port it runs on. If you don’t know, you can find the ports that a container image exposes with this command

$ docker inspect <image-name>:<image-tag> -f '{{range $port, $val := .ContainerConfig.ExposedPorts}}{{$port}} {{end}}'

Note

Replace <image-name>:<image-tag> with the your Docker image

It’s possible that multiple or no ports are listed. If that’s the case you’ll need to refer to the documentation for the tool to figure out the port it uses. Make a note of the port, you’ll need it later when integrating the tool in DE. Here are the tools and their ports for common tools such as Jupyter notebook, Rstudio and Shiny. If you are developing any applications based on these tools, you can use these ports while integrating the tool in DE.

Type

Port

Jupyter

8888

Rstudio

80

Shiny

3838

2. Get the UID of the tool’s user

You’ll need to figure out the UID of the of the user the app runs as. Many tools will start up as root and then use another user for the actual process, so it might take a little investigation to figure this out. To start this figure out the user that the container is configured to start up using:

$ docker inspect <image-name>:<image-tag> -f '{{.ContainerConfig.User}}'

If you’re lucky that will contain the numerical UID of the user. In that case you can make a note of the UID and move on. Otherwise you have more work to do. The User field can also be empty or set to the username. If its empty, then the user is root. If it’s a username then you’ll need to get the UID from inside the container.

To get the UID for a username run this:

$ docker run --rm -it --entrypoint "id" <image-name>:<image-tag> -u <username>

Note

Replace <image-name>:<image-tag> with the your Docker image

If the User field is empty or root, you need to be sure that the process inside the container actually runs as root. There are a few ways to check this:

  • Fire up the container, exec into it, and do a ps aux to see the user the process is running as.

$ docker run -d --name app <image-name>:<image-tag>

$ docker exec -it app ps aux

  • Print out the contents of /etc/passwd and check for hints:

docker run --rm -it --entrypoint "cat" <image-name>:<image-tag> /etc/passwd

  • Alternatively check the documentation for the tool.

Note

The UID of the tool can be empty but setting the UID will make sure that the user can write to the input files in the container.

Make a note of the UID, you’ll need it later when putting together the JSON for the tool and app.

3. Get the working directory

You’ll need the working directory for the process in the tool container, which may not correspond to the default working directory for the container.

To get the default working directory for the container run this:

$ docker inspect <image-name>:<image-tag> -f '{{.ContainerConfig.WorkingDir}}'

  • If that prints out an empty string, then the default working directory is /.

  • If the container fires up as root but the tool runs as another user, then the working directory may need to be that user’s home directory.

  • If the container changes to another directory after it starts up, then the working directory may need to be that directory.

  • If all else fails, check the documentation and/or try out the container locally to figure out what it does.

Important

Keep in mind that the working directory is where the input files will be made available. Similar to UID, working directory is not mandatory but given jupyter lab’s default behavior of showing things in subdirectories of the place it’s started. So if you’re loading notebooks and data from the Datastore, you want the working directory (where those files are loaded into the container) to be in the right spot

Make a note of the working directory, you’ll need it later when putting together the JSON for the tool and app.

CyVerse_logo

Home_Icon Learning Center Home

Contact

Intercom

Ask Forum

Fix or improve this documentation:

Fix or improve this documentation