This is the multi-page printable view of this section. Click here to print.

Return to the regular view of this page.

Tutorials

1: Customize your shell
2: Apptainer tutorial

1 - Customize your shell

After logging in to DAIC you can customize your shell.

Source `.bashrc` upon login

.bashrc is a configuration file for the Bash shell, which is the default command-line shell on many Linux and Unix-based systems. It is a hidden file located in the user’s home directory (~/.bashrc) and is executed every time a new interactive Bash session starts. The file contains settings that customize the shell’s behavior, such as defining environment variables, setting prompt appearance, and specifying terminal options.

Edit or create the file ~/.profile and insert the following line:

source ~/.bashrc

Now, your .bashrc will be loaded upon login.

Customize prompt and aliases.

Aliases are custom shortcuts or abbreviations for longer commands in the shell. They allow you to define a shorter, user-friendly command that, when executed, will perform a longer or more complex command. For example, you might create an alias like alias ll='ls -la' in your .bashrc file to quickly list all files in a directory in long format. Aliases can help improve productivity by saving time and effort when working with the shell. You can set these configurations permanently by editing your ~/.bashrc file.

## Add these lines to your ~/.bashrc file to make use of these settings.

# Alias
alias ll='ls -alF'
alias la='ls -A'
alias ls='ls --color=auto'
alias l='ls -rtlh --full-time --color=auto'
alias md='mkdir'
alias ..='cd ..'
alias ...='cd ../..'
alias ....='cd ../../..'
alias src='source ~/.bashrc'

## Slurm helpers
alias interactive='srun --pty --nodes=1 --ntasks=1 --cpus-per-task=4 --mem=8G --time=1:00:00 bash'
alias st='sacct --format=JobID,JobName%30,State,Elapsed,Timelimit,AllocNodes,Priority,Start,NodeList'
alias sq="squeue -u $USER --format='%.18i %.12P %.30j %.15u %.2t %.12M %.6D %R'"
alias slurm-show-my-accounts='sacctmgr list user "$USER" withassoc format="user%-20,account%-45,maxjobs,maxsubmit,maxwall,maxtresperjob%-40"'
alias slurm-show-all-accounts='sacctmgr show account format=Account%30,Organization%30,Description%60'
alias slurm-show-nodes='sinfo -lNe'

# Shellstyle

## Assuming your shell background is black!

## Prompt setting (readable prompt colors and username@hostname:)
export PS1='\[\033[01;96m\]\u\[\033[0m\]@\[\033[01;32m\]\h\[\033[0m\]:\[\033[96m\]\w\[\033[00m\]\$ '

## ls colors (readable ls colors)
export LS_COLORS='rs=0:di=1;35:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:su=37;41:sg=30;43:ca=30;41:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arj=01;31:*.taz=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lz=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.webm=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=00;36:*.au=00;36:*.flac=00;36:*.mid=00;36:*.midi=00;36:*.mka=00;36:*.mp3=00;36:*.mpc=00;36:*.ogg=00;36:*.ra=00;36:*.wav=00;36:*.axa=00;36:*.oga=00;36:*.spx=00;36:*.xspf=00;36:';

REIT bash configuration

An example configuration is available here: https://gitlab.ewi.tudelft.nl/reit/shell-config

2 - Apptainer tutorial

Using Apptainer to containerize environments.

What is containerization?

Imagine you want to move your belongings from one place to another. You could just pile everything into a truck, but things might shift around, break, or get mixed up along the way. Instead, you might pack your stuff into separate boxes: one box for clothes, one for kitchen items, one for books, and so on. This way, everything is organized and protected, and you can easily move the boxes around.

Containerization in computing works similarly. When you want to run software or applications, you can pack them into “containers” rather than just running them directly on your computer. These containers are like those boxes—they contain everything the application needs to run, such as code, libraries, and settings. This makes the application portable and consistent.

Why it’s helpful?

Consistency: Because the application runs inside a container, it behaves the same way regardless of where it’s running. This means you can develop on one computer, test on another, and deploy on a server without worrying about differences between environments.
Isolation: Each container is independent from others. This keeps applications from interfering with each other or with the host system, enhancing security and reliability.
Portability: Containers can run on different machines without modification, making it easier to move applications from one server to another, or even from a local computer to the cloud.
Efficiency: Containers share the host system’s resources like the operating system, which makes them lightweight and fast to start up compared to virtual machines.

On DAIC specifically, many users encounter issues with limited home directory sizes and Windows-based /tudelft.net mounts (See Storage), which can hinder the use of conda/mamba and/or pip due to compatibility challenges. Containers offer a solution by enabling users to encapsulate their software and dependencies in a portable, self-contained environment. This means users can store a container e.g. on the staff-umbrella storage with all necessary dependencies, including those installed with pip. This enables users to create and use multiple large environments and run applications reliably and reproducibly, without running into limitations from Windows-based mounts or small home directories.

Containerization technology (Apptainer)

Containerization is a convenient means to deploy libraries and applications to different environments in a reproducible manner. DAIC supports Apptainer (previously Singularity), an open-source container platform, designed to run complex applications on HPC clusters. Apptainer makes it possible to use docker images natively at a higher level of security and isolation. A container image, typically a *.sif file, is a self-contained file with all necessary components to run an application, including code, runtime libraries, and dependencies.

The definition file (*.def) contains the recipe to build an image.
An image (*.sif) is a complete package that includes everything needed to run an application, such as code, libraries, and settings. It only needs Apptainer to be run.
A container is a running instance of an image with its own working space, so it can hold changes and temporary data such as ongoing calculations as you interact with the application. This could mean training a machine learning model for example.

How to run commands/programs inside a container?

Generally, to launch a container image, your commands look as follows:

$ apptainer shell <container> # OR
$ apptainer exec  <container> <command>
$ apptainer run   <container>

where:

<container> is the path to a container image, typically, a *.sif file
<command> is the command you like to run from inside the container, eg, hostname
Both shell and exec can be used to launch container images. The difference is that shell allows you to work inside the container image interactively; while exec executes the <command> inside the image and exits. Of course, by using something like /bin/bash as the <command>, exec behaves exactly like shell.
run also launches a container image, but runs the default action defined in the container image. See an example use case in Building images

The question is now: where to get the <container> file from? You can either:

use a pre-built image by pulling from a repository (see Pulling images), or,
build your own container image and use it accordingly (see Building images).

Note

If you intend to extensively work/test your image interactively, it is best to first submit an interactive SLURM job with the needed resources, eg, memory, gpus, … etc:

$ hostname  # To check this is DAIC. login[1-3] are the login nodes
login1.daic.tudelft.nl 
$ sinteractive # Default resources: --time=01:00:00 --cpus-per-task = 2 --mem=1024 
Note: interactive sessions are automatically terminated when they reach their time limit (1 hour)!
srun: job 8543393 queued and waiting for resources
srun: job 8543393 has been allocated resources
 13:35:30 up 5 days,  3:41,  0 users,  load average: 8,79, 7,60, 7,11
$ hostname # To check we are on a compute node
grs3.daic.tudelft.nl

How to get container files?

Pulling images

Many repositories exist where container images are hosted. Apptainer allows pulling and using images from repositories like DockerHub, BioContainers and NVIDIA GPU Cloud (NGC).

Pulling from DockerHub

For example, to obtain the latest Ubuntu image from DockerHub:

$ hostname # check this is DAIC
login1.daic.tudelft.nl
$ cd && mkdir containers && cd containers # as convenience, use this directory
$ apptainer pull docker://ubuntu:latest # actually pull the image
INFO:    Converting OCI blobs to SIF format
INFO:    Starting build...
Getting image source signatures
Copying blob 837dd4791cdc done  
Copying config 1f6ddc1b25 done  
Writing manifest to image destination
Storing signatures
...
INFO:    Creating SIF file...

Now, to check the obtained image file:

$ ls  
ubuntu_latest.sif
$ apptainer exec ubuntu_latest.sif cat /etc/os-release # execute cat command and exit
PRETTY_NAME="Ubuntu 22.04.2 LTS"
NAME="Ubuntu"
VERSION_ID="22.04"
VERSION="22.04.2 LTS (Jammy Jellyfish)"
VERSION_CODENAME=jammy
ID=ubuntu
ID_LIKE=debian
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
UBUNTU_CODENAME=jammy

$ ls /.singularity.d/ # container-specific directory should not be found on host
ls: cannot access /.singularity.d/: No such file or directory

$ apptainer shell ubuntu_latest.sif # launch container interactively
Apptainer>
Apptainer> hostname
login1.daic.tudelft.nl
Apptainer> ls
ubuntu_latest.sif
Apptainer> ls /.singularity.d/ 
Singularity  actions  env  labels.json  libs  runscript  startscript
Apptainer> exit

In the above snippet, note:

The command prompt changes within the container to Apptainer>
The container seamlessly interacts with the host system. For example, it inherits its hostname (the DAIC login node in this case). The container also inherits the $HOME variable, and is able to edit/delete files from there.
The container has its own file system, which is distinct from the host. The presence of a directory like /.singularity.d is another feature of the specific to the container.

Warning

To isolate files in your system (ie, your local machine or DAIC) from the files inside the container (and thus, avoid possible erroneous deletes/edits), it is recommended to add a -c or -C flags to your apptainer commands

$ apptainer shell -C ubuntu_latest.sif

Pulling from NVIDIA GPU cloud (NGC)

This is a specialized registry provided by NVIDIA for GPU accelerated applications or GPU software development tools. These images are large, and one is recommended to download them locally in your machine, and only send the downloaded image to DAIC. For this, you need to have Apptainer locally installed first. To install Apptainer in your machine, follow the official Installing Apptainer instructions. Apptainer needs a Linux kernel to run, if you create your container on a MacBook, or a computer with a different CPU architecture than the target system, there is a good chance that the container will not run.

Warning

By default, Apptainer images are saved to ~/.singularity. Ideally, to avoid quota issues, you’d set the environment variable SINGULARITY_CACHEDIR to a different location. At present, both the bulk and umbrella filesystems do not support pulling images, so you are advised to pull these to your local machine and then copy over the image file to DAIC.

$ hostname #check this is your own PC/laptop
$ apptainer pull docker://nvcr.io/nvidia/pytorch:23.05-py3
$ scp pytorch_23.05-py3.sif  hpc-login:/tudelft.net/staff-umbrella/...<YourDirectory>/apptainer

Now, to check this particular image on DAIC:

$ hostname # check this is DAIC not your own PC/laptop
login1.daic.tudelft.nl
$ cd /tudelft.net/staff-umbrella/...<YourDirectory>/apptainer # path where you put images
$ apptainer shell -C --nv pytorch_23.05-py3.sif  #--nv to use NVIDIA GPU and have CUDA support
Apptainer>
Apptainer> hostname
login1.daic.tudelft.nl # hostname inherited
Apptainer> ls /.singularity.d/ # verify this is the image
Singularity  actions  env  labels.json  libs  runscript  startscript

Building images

If you prefer (or need) to have a custom container image, then you can build your own container image from a definition file, typically *.def file, that sets up the image with your custom dependencies. The only requirement for building is to be in a machine (eg, your local laptop/pc) where you have sudo/root privileges. In other words, you can not build images on DAIC directly: First, you should build the image locally, and then send it to DAIC to run there.

An example definion file, cuda_based.def, for a cuda-enabled container may look as follows:

$ cat cuda_based.def
# Header
Bootstrap: docker
From: nvidia/cuda:12.1.1-devel-ubuntu22.04

# (Optional) Sections/ data blobs
%post
    apt-get update # update system
    apt-get install -y git   # install git
    git clone https://github.com/NVIDIA/cuda-samples.git  # clone target repository
    cd cuda-samples
    git fetch origin --tags && git checkout v12.1 # fetch certain repository version
    cd Samples/1_Utilities/deviceQuery && make # install certain tool

%runscript
    /cuda-samples/Samples/1_Utilities/deviceQuery/deviceQuery

where:

The header, the first 2 lines of this example, specify the source of a base image, (eg, Bootstrap: docker), and the base image (From: nvidia/cuda:12.1.1-devel-ubuntu22.04) to be pulled from this source. The container image will be built on top of this base image. In this example, the base image will be built from Ubuntu 22.04 OS with the CUDA toolkit 12.1 pre-installed.
The rest of the file are optional data blobs or sections. In this example, the following blobs are used:
- %post blob: the steps to download, configure and install needed custom software and libraries on the base image. In this example, the steps install git, clone a repo, and install a package via make
- %runscript blob: the scripts or commands to execute when the container image is run. That is, this code is the entry point to the container with the run command. In this example, the deviceQuery is executed once the container is run.
- Other blobs may be present in the def file. See Definition files documentation for more details and examples.

And now, build this image and send it over to DAIC:

$ hostname #check this is your machine
$ sudo apptainer build cuda_based_image.sif cuda_based.def # building may take ~ 2-5 min, depending on your internet
INFO:    Starting build...
Getting image source signatures
Copying blob d5d706ce7b29 [=>------------------------------------] 29.2MiB / 702.5MiB
...
INFO:    Adding runscript
INFO:    Creating SIF file...
INFO:    Build complete: cuda_based_image.sif  
$
$ scp cuda_based_image.sif  hpc-login:/tudelft.net/staff-umbrella/...<YourDirectory>/apptainer # send to DAIC

On DAIC, check the image:

$ hostname # check you are on DAIC
login1.daic.tudelft.nl 
$ sinteractive --cpus-per-task=2 --mem=1024 --gres=gpu --time=00:05:00 # request a gpu node
$ hostname # check you are on a compute node
insy13.daic.tudelft.nl

$ apptainer run --nv -C cuda_based_image.sif # --nv to use NVIDIA GPU and have CUDA support
/cuda-samples/Samples/1_Utilities/deviceQuery/deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "NVIDIA GeForce GTX 1080 Ti"
  CUDA Driver Version / Runtime Version          12.1 / 12.1
  CUDA Capability Major/Minor version number:    6.1
  Total amount of global memory:                 11172 MBytes (11714887680 bytes)
  (028) Multiprocessors, (128) CUDA Cores/MP:    3584 CUDA Cores
  GPU Max Clock rate:                            1582 MHz (1.58 GHz)
  Memory Clock rate:                             5505 Mhz
  Memory Bus Width:                              352-bit
  L2 Cache Size:                                 2883584 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
  Maximum Layered 1D Texture Size, (num) layers  1D=(32768), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(32768, 32768), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total shared memory per multiprocessor:        98304 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 2 copy engine(s)
  Run time limit on kernels:                     No
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device supports Managed Memory:                Yes
  Device supports Compute Preemption:            Yes
  Supports Cooperative Kernel Launch:            Yes
  Supports MultiDevice Co-op Kernel Launch:      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 141 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.1, CUDA Runtime Version = 12.1, NumDevs = 1
Result = PASS

Warning

Always pass --nv to apptainer to run GPU-accelerated applications or libraries inside the container. Note that you also need 1) your host system must have NVIDIA GPU drivers installed and compatible with the version of Singularity you are using, and 2) the container you are running should have the necessary dependencies and configurations to support GPU acceleration.

$ apptainer shell --nv -C cuda_based_image.sif

Note

Building container images from a definition file is recommended to ensure the reproducibility of the resulting container image. However, there can be cases of complex dependencies where it is not clear upfront how the software installations and dependencies should be set up. In such cases, it is possible to interactively develop the image by building it in writable sandbox mode first. In such cases, take note of all installation commands used in the sandbox, so you can include them in a recipe file. See Apptainer Sandbox Directories for more details.

Extending existing images

During software development, it is common to incrementally build code and go through many iterations of debugging and testing. A development container may be used in this process. In such scenarios, re-building the container from the base image with each debugging or testing iteration becomes taxing very quickly, due to dependencies and installations involved. Instead, the Bootstrap: localimage and From:<path/to/local/image> header can be used to base the development container on some local image.

As an example, assume it is desirable to develop some code on the basis of the cuda_based.sif image created in the Building images section. Building from the original cuda_based.def file can take ~ 4 minutes. However, if the *.sif file is already available, building on top of it, via a dev_on_cuda_based.def file as below, takes ~ 2 minutes. This is already a time saving factor of 2 in this case.

$ hostname # check this is your machine
$ cat dev_on_cuda_based.def # def file for an image based on localimage
# Header
Bootstrap: localimage
From: cuda_based.sif

# (Optional) Sections/ data blobs
%runscript
    echo "Arguments received: $*"
    exec echo "$@"

$
$ sudo apptainer build dev_image.sif # build the image
INFO:    Starting build...
INFO:    Verifying bootstrap image cuda_based.sif
WARNING: integrity: signature not found for object group 1
WARNING: Bootstrap image could not be verified, but build will continue.
INFO:    Adding runscript
INFO:    Creating SIF file...
INFO:    Build complete: dev_image.sif
$
$ apptainer run  dev_image.sif "hello world" # check runscript of the new def file is executed
INFO:    gocryptfs not found, will not be able to use gocryptfs
Arguments received: hello world
hello world
$
$ apptainer shell  dev_image.sif # further look inside the image
Apptainer> 
Apptainer> ls /cuda-samples/Samples/1_Utilities/deviceQuery/deviceQuery # commands installed in the original image are available
/cuda-samples/Samples/1_Utilities/deviceQuery/deviceQuery

Apptainer>
Apptainer> cat /.singularity.d/bootstrap_history/Apptainer0 # The original def file is also preserved 
bootstrap: docker
from: nvidia/cuda:12.1.1-devel-ubuntu22.04

%runscript
    /cuda-samples/Samples/1_Utilities/deviceQuery/deviceQuery  


%post
    apt-get update # update system
    apt-get install -y git   # install git
    git clone https://github.com/NVIDIA/cuda-samples.git  # clone target repository
    cd cuda-samples
    git fetch origin --tags && git checkout v12.1 # fetch certain repository version
    cd Samples/1_Utilities/deviceQuery && make # install certain tool

As can be seen in this example, the new def file not only preserves the dependencies of the original image, but it also preserves a complete history of all build processes while giving flexible environment that can be customized as need arises.

Deploying conda and pip in a container

There might be situations where you have a certain conda environment in your local machine that you need to set up in DAIC to commence your analysis. In such cases, deploying your conda environment in a container and sending this container to DAIC does the job for you.

As an example, let’s create a simple demo environment, environment.yml in our local machine,

name: __apptainer-env__
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.9
  - matplotlib
  - pip
  - pip:
    - -r requirements.txt

And everything that should be installed with pip in requirement.txt file:

--extra-index-url https://download.pytorch.org/whl/cu123
torch
annoy

Now, it is time to create the container definition file. One option is to base the image on condaforge/miniforge, which is a minimal Ubuntu installation with conda preinstalled at /opt/conda:

Bootstrap: docker
From: condaforge/miniforge3:latest

%files
    environment.yml /environment.yml
    requirements.txt /requirements.txt

%post
    # Update and install necessary packages
    apt-get update && apt-get install -y tree time vim ncdu speedtest-cli build-essential

    # Create a new Conda environment using the environment files.
    mamba env create --quiet --file /environment.yml
    
    # Clean up
    apt-get clean && rm -rf /var/lib/apt/lists/*
    mamba clean --all -y

    # Now add the script to activate the Conda environment
    echo '. "/opt/conda/etc/profile.d/conda.sh"' >> $SINGULARITY_ENVIRONMENT
    echo 'conda activate __apptainer-env__' >> $SINGULARITY_ENVIRONMENT

This file is similar to the file in the Building images, with the addition of %files area. %files specifies the files in the host system (ie, your machine) that need to be copied to the container image, and optionally, where should they be available. In the previous example, the environment.yml file will be available in /opt/ in the container.

Now, time to build and check the image:

$ apptainer build demo-env-image.sif demo-env-recipe.def
INFO:    Starting build...
Getting image source signatures
...
INFO:    Creating SIF file...       
INFO:    Build complete: demo-env-image.sif
...

We are going to use the environment inside a container together with a Python script that we store outside the container. Create the file analysis.py, which generate a plot:

#!/usr/bin/env python3

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 2 * np.pi, 100)
y = np.sin(x)

plt.plot(x, y)
plt.title('Sine Wave')
plt.savefig('sine_wave.png')

Now, we can run the analysis:

$ apptainer exec demo-env-image.sif python analysis.py  
$ ls # check the image file was created
sine_wave.png

Warning

In the last example, the container read and wrote a file to the host system directly. This behavior is risky. You are strongly recommended to expose only the desired host directories to the container. See Exposing host directories

Exposing host directories

Depending on use case, it may be necessary for the container to read or write data from or to the host system. For example, to expose only files in a host directory called ProjectDataDir to the container image’s /mnt directory, add the --bind directive with appropriate <hostDir>:<containerDir> mapping to the commands you use to launch the container, in conjunction with the -C flag eg, shell or exec as below:

$ ls  # check ProjectDataDir exists
$ ls ProjectDataDir # check contents of ProjectDataDir
raw_data.txt
$ apptainer shell -C --bind ProjectDataDir:/mnt ubuntu_latest.sif # Launch the container with ProjectDataDir bound
Apptainer> ls
Apptainer> ls /mnt # check the files are accessible inside the container
raw_data.txt
Apptainer> echo "Date: $(date)" >> raw_data.txt # edit the file
Apptainer> tail -n1 raw_data.txt # check the date was written to the file
Apptainer> exit # exit the container
$ tail -n1 ProjectDataDir/raw_data.txt # check the change persisted

If the desire is to expose this directory as read-only inside the container, the --mount directive should be used instead of --bind, with rodesignation as follows:

apptainer shell -C --mount type=bind,source=ProjectDataDir,destination=/mnt,ro ubuntu_latest.sif # Launch the container with ProjectDataDir bound
Apptainer> ls /mnt # check the files are accessible inside the container
raw_data.txt
Apptainer> echo "Date: $(date)" >> /mnt/raw_data.txt # attempt to edit fails
bash: tst: Read-only file system

Advanced: containers and (fake) native installation

It’s possible to use Apptainer to install and then use software as if it were installed natively in the host system. For example, if you are a bioinformatician, you may be using software like samtools or bcftools for many of your analyses, and it may be advantageous to call it directly. Let’s take this as an illustrative example:

For hygiene, create the following file hierarchy: below a software directory an exec directory to put the container images and other executables, and a bin directory to contain softlinks:

$ mkdir -p software/bin/ software/exec

Create the image definition file (or pull from a repository, as appropriate) and build:

$ cd software/exec
$
$ cat bio-recipe.def
Bootstrap: docker
From: ubuntu:latest
%post
    apt-get update                       # update system
    apt-get install -y samtools bcftools # install software
    apt-get clean                        # clean up

$ sudo apptainer build bio-container.sif bio-recipe.def

Now, create the following wrapper script:

$ cat wrapper_bio-container.sh
#!/bin/bash
containerdir="$(dirname $(readlink -f ${BASH_SOURCE[0]}))"
cmd="$(basename $0)"
apptainer exec "${containerdir}/bio-container.sif" "$cmd" "$@"
$
$ chmod +x wrapper_bio-container.sh # make it executable

Create the softlinks:

$ cd ../bin
$ ln -s ../exec/wrapper_bio-container.sh  samtools
$ ln -s ../exec/wrapper_bio-container.sh  bcftools

Add the installation directory to your $PATH variable, and you will be able to call these tools

$ export PATH=$PATH:$PWD
$
$ bcftools -v
INFO:    gocryptfs not found, will not be able to use gocryptfs
bcftools 1.13
Using htslib 1.13+ds
Copyright (C) 2021 Genome Research Ltd.
License Expat: The MIT/Expat license
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
$
$ samtools version                                  
INFO:    gocryptfs not found, will not be able to use gocryptfs                             
samtools 1.13                                                                               
Using htslib 1.13+ds                                                                        
Copyright (C) 2021 Genome Research Ltd.

Note

At the end of the previous steps, you will get the following tree structure. Please be mindful of when and where commands were executed.

$ tree software/
software/
├── bin
│   ├── bcftools -> ../exec/wrapper.sh
│   └── samtools -> ../exec/wrapper.sh
└── exec
    ├── bio-container.sif
    └── wrapper.sh

To permanently reflect changes to your $PATH variable, you may wish to add the step:

echo export PATH=$PATH:$PWD >> ~/.bash_profile

Tutorials

1 - Customize your shell

Source .bashrc upon login

Customize prompt and aliases.

REIT bash configuration

2 - Apptainer tutorial

What is containerization?

Why it’s helpful?

Containerization technology (Apptainer)

How to run commands/programs inside a container?

Note

How to get container files?

Pulling images

Pulling from DockerHub

Warning

Pulling from NVIDIA GPU cloud (NGC)

Warning

Building images

Warning

Note

Extending existing images

Deploying conda and pip in a container

Warning

Exposing host directories

Advanced: containers and (fake) native installation

Note

Source `.bashrc` upon login