Reproducible computational environments using containers: Introduction to Singularity

Introducing Containers

Overview

Teaching: 20 min
Exercises: 0 min
Questions

  • What are containers, and why might they be useful to me?

Objectives

  • Show how software depending on other software leads to configuration management problems.

  • Identify the problems that software installation can pose for research.

  • Explain the advantages of containerization.

  • Explain how using containers can solve software configuration problems

Learning about Containers

The Australian Research Data Commons has produced a short introductory video about containers that covers many of the points below. Watch it before or after you go through this section to reinforce your understanding!

How can software containers help your research?

Australian Research Data Commons, 2021. How can software containers help your research?. [video] Available at: https://www.youtube.com/watch?v=HelrQnm3v4g DOI: http://doi.org/10.5281/zenodo.5091260

Scientific Software Challenges

What’s Your Experience?

Take a minute to think about challenges that you have experienced in using scientific software (or software in general!) for your research. Then, share with your neighbors and try to come up with a list of common gripes or challenges.

You may have come up with some of the following:

A lot of these characteristics boil down to one fact: the main program you want to use likely depends on many, many, different other programs (including the operating system!), creating a very complex, and often fragile system. One change or missing piece may stop the whole thing from working or break something that was already running. It’s no surprise that this situation is sometimes informally termed “dependency hell”.

Software and Science

Again, take a minute to think about how the software challenges we’ve discussed could impact (or have impacted!) the quality of your work. Share your thoughts with your neighbors. What can go wrong if our software doesn’t work?

Unsurprisingly, software installation and configuration challenges can have negative consequences for research:

Thankfully there are ways to get underneath (a lot of) this mess: containers to the rescue! Containers provide a way to package up software dependencies and access to resources such as files and communications networks in a uniform manner.

What is a Container?

To understand containers, let’s first talk briefly about your computer.

Your computer has some standard pieces that allow it to work – often what’s called the hardware. One of these pieces is the CPU or processor; another is the amount of memory or RAM that your computer can use to store information temporarily while running programs; another is the hard drive, which can store information over the long-term. All these pieces work together to do the “computing” of a computer, but we don’t see them because they’re hidden from view (usually).

Instead, what we see is our desktop, program windows, different folders, and files. These all live in what’s called the filesystem. Everything on your computer – programs, pictures, documents, the operating system itself – lives somewhere in the filesystem.

NOW, imagine you want to install some new software but don’t want to take the chance of making a mess of your existing system by installing a bunch of additional stuff (libraries/dependencies/etc.). You don’t want to buy a whole new computer because it’s too expensive. What if, instead, you could have another independent filesystem and running operating system that you could access from your main computer, and that is actually stored within this existing computer?

Or, imagine you have two tools you want to use in your groundbreaking research on cat memes: PurrLOLing, a tool that does AMAZINGLY well at predicting the best text for a meme based on the cat species and WhiskerSpot, the only tool available for identifying cat species from images. You want to send cat pictures to WhiskerSpot, and then send the species output to PurrLOLing. But there’s a problem: PurrLOLing only works on Ubuntu and WhiskerSpot is only supported for OpenSUSE so you can’t have them on the same system! Again, we really want another filesystem (or two) on our computer that we could use to chain together WhiskerSpot and PurrLOLing in a “pipeline”…

Container systems, like Singularity/Apptainer and Docker, are special programs on your computer that make it possible! The term “container” can be usefully considered with reference to shipping containers. Before shipping containers were developed, packing and unpacking cargo ships was time consuming and error prone, with high potential for different clients’ goods to become mixed up. Just like shipping containers keep things together that should stay together, software containers standardize the description and creation of a complete software system: you can drop a container into any computer with the container software installed (the ‘container host’), and it should “just work”.

Virtualization

Containers are an example of what’s called virtualization – having a second “virtual” computer running and accessible from a main or host computer. Another example of virtualization are virtual machines or VMs. A virtual machine typically contains a whole copy of an operating system in addition to its own filesystem and has to get booted up in the same way a computer would. A container is considered a lightweight version of a virtual machine; underneath, the container is (usually) using the Linux kernel and simply has some flavour of Linux + the filesystem inside.

One final term: while the container is an alternative filesystem layer that you can access and run from your computer, the container image is the ‘recipe’ or template for a container. The container image has all the required information to start up a running copy of the container. A running container tends to be transient and can be started and shut down. The container image is more long-lived, as a definition for the container. You could think of the container image like a cookie cutter – it can be used to create multiple copies of the same shape (or container) and is relatively unchanging, where cookies come and go. If you want a different type of container (cookie) you need a different container image (cookie cutter).

Putting the Pieces Together

Think back to some of the challenges we described at the beginning. The many layers of scientific software installations make it hard to install and re-install scientific software – which ultimately, hinders reliability and reproducibility.

But now, think about what a container is – a self-contained, complete, separate computer filesystem. What advantages are there if you put your scientific software tools into containers?

This solves several of our problems:

The rest of this workshop will show you how to download and run containers from pre-existing container images on your own computer, and how to create and share your own container images.

Use cases for containers

Now that we have discussed a little bit about containers – what they do and the issues they attempt to address – you may be able to think of a few potential use cases in your area of work. Some examples of common use cases for containers in a research context include:

Key Points

  • Almost all software depends on other software components to function, but these components have independent evolutionary paths.

  • Small environments that contain only the software that is needed for a given task are easier to replicate and maintain.

  • Critical systems that cannot be upgraded, due to cost, difficulty, etc. need to be reproduced on newer systems in a maintainable and self-documented way.

  • Virtualization allows multiple environments to run on a single computer.

  • Containerization improves upon the virtualization of whole computers by allowing efficient management of the host computer’s memory and storage resources.

  • Containers are built from ‘recipes’ that define the required set of software components and the instructions necessary to build/install them within a container image.

  • Singularity and Docker are examples of software platforms that can create containers and the resources they use.

Singularity: Getting started

Overview

Teaching: 15 min
Exercises: 10 min
Questions

  • What is Singularity and why might I want to use it?

Objectives

  • Understand what Singularity is and when you might want to use it.

  • Undertake your first run of a simple Singularity container.

The episodes in this lesson will introduce you to the Singularity container platform and demonstrate how to set up and use Singularity.

What is Singularity?

Singularity (or Apptainer, we’ll get to this in a minute…) is a container platform that supports packaging and deploying software and tools in a portable and reproducible manner.

You may be familiar with Docker, another container platform that is now used widely. If you are, you will see that in some ways, Singularity is similar to Docker. However, in other ways, particularly in terms of the system’s architecture, it is fundamentally different. These differences mean that Singularity is particularly well-suited to running on shared platforms such as distributed, High Performance Computing (HPC) infrastructure, as well as on a Linux laptop or desktop.

Singularity runs containers from container images which, as we discussed, are essentially a virtual computer disk that contains all of the necessary software, libraries and configuration to run one or more applications or undertake a particular task, e.g. to support a specific research project. This saves you the time and effort of installing and configuring software on your own system or setting up a new computer from scratch, as you can simply run a Singularity container from an image and have a virtual environment that is equivalent to the one used by the person who created the image. Singularity/Apptainer is increasingly widely used in the research community for supporting research projects due to its support for shared computing platforms.

System administrators will not, generally, install Docker on shared computing platforms such as lab desktops, research clusters or HPC platforms because the design of Docker presents potential security issues for shared platforms with multiple users. Singularity/Apptainer, on the other hand, can be run by end-users entirely within “user space”, that is, no special administrative privileges need to be assigned to a user in order for them to run and interact with containers on a platform where Singularity has been installed.

A little history…

Singularity is open source software and was initially developed within the research community. A couple of years ago, the project was “forked” something that is not uncommon within the open source software community, with the software effectively splitting into two projects going in different directions. The fork is being developed by a commercial entity, Sylabs.io who provide both the free, open source SingularityCE (Community Edition) and Pro/Enterprise editions of the software. The original open source Singularity project has recently been renamed to Apptainer and has moved into the Linux Foundation. The initial release of Apptainer was made about a year ago, at the time of writing. While earlier versions of this course focused on versions of Singularity released before the project fork, we now base the course material on recent Apptainer releases. Despite this, the basic features of Apptainer/Singularity remain the same and so this material is equally applicable whether you’re working with a recent Apptainer release or a slightly older Singularity version. Nonetheless, it is useful to be aware of this history and that you may see both Singularity and Apptainer being used within the research community over the coming months and years.

Another point to note is that some systems that have a recent Apptainer release installed may also provide a singularity command that is simply a link to the apptainer executable on the system. This helps to ensure that existing scripts being used on the system that were developed before the migration to Apptainer will still function correctly.

For now, the remainder of this material refers to Singularity but where you have a release of Apptainer installed on your local system, you can simply replace references to singularity with apptainer, if you wish.

Checking Singularity works

Login to ARCHER2 using the login address training.dyn.archer2.ac.uk:

ssh -i /path/to/ssh-key user@training.dyn.archer2.ac.uk

Now check that the singularity command is available in your terminal:

remote$ singularity --version

singularity version 3.7.3-1

Loading a module

HPC systems often use modules to provide access to software on the system so you may need to use the command:

remote$ module load singularity

before you can use the singularity command on remote systems. However, this depends on how the system is configured. You do not need to load a module on ARCHER2. If in doubt, consult the documentation for the system you are using or contact the support team.

Images and containers: reminder

A quick reminder on terminology: we refer to both container images and containers. What is the difference between these two terms?

Container images (sometimes just images) are bundles of files including an operating system, software and potentially data and other application-related files. They may sometimes be referred to as a disk image or image and they may be stored in different ways, perhaps as a single file, or as a group of files. Either way, we refer to this file, or collection of files, as an image.

A container is a virtual environment that is based on a container image. That is, the files, applications, tools, etc that are available within a running container are determined by the image that the container is started from. It may be possible to start multiple container instances from an image. You could, perhaps, consider an image to be a form of template from which running container instances can be started.

Getting a container image and running a Singularity container

Singularity uses the Singularity Image Format (SIF) and container images are provided as single SIF files (usually with a .sif or .img filename extension). Singularity container images can be pulled from the Sylabs Cloud Library, a registry for Singularity container images. Singularity is also capable of running containers based on container images pulled from Docker Hub and other Docker image repositories (e.g. Quay.io). We will look at accessing container images from Docker Hub later in the course.

Sylabs Remote Builder

Note that in addition to providing a repository that you can pull container images from, Sylabs Cloud Library can also build Singularity images for you from a recipe - a configuration file defining the steps to build an image. We will look at recipes and building images later in the workshop.

Pulling a container image from Sylabs Cloud Library

Let’s begin by creating a test directory, changing into it and pulling an existing container image from Sylabs Cloud Library:

remote$ mkdir test
remote$ cd test
remote$ singularity pull lolcow.sif library://lolcow

INFO:    Downloading library image
 90.4 MiB / 90.4 MiB [===============================================================================================================] 100.00% 90.4 MiB/s 1s

What just happened? We pulled a container image from a remote repository using the singularity pull command and directed it to store the container image in a file using the name lolcow.sif in the current directory. If you run the ls command, you should see that the lolcow.sif file is now present in the current directory.

remote$ ls -lh

total 60M
-rwxr-xr-x. 1 auser group 91M Jun 13  2023 lolcow.sif

Running a Singularity container

We can now run a container based on the lolcow.sif container image:

remote$ singularity run lolcow.sif

 ______________________________
< Tue Jun 20 08:44:51 UTC 2023 >
 ------------------------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||

The above command ran a lolcow container based on the container image we downloaded from the online repository and the resulting output was shown.

What just happened? When we use the singularity run command, Singularity does three things:

1. Starts a Running Container 2. Performs Default Action 3. Shuts Down the Container
Starts a running container, based on the container image. Think of this as the “alive” or “inflated” version of the container – it’s actually doing something. If the container has a default action set, it will perform that default action. This could be as simple as printing a message (as above) or running a whole analysis pipeline! Once the default action is complete, the container stops running (or exits).

Default action

How did the container determine what to do when we ran it? What did running the container actually do to result in the displayed output?

When you run a container from a Singularity container image using the singularity run command, the container runs the default run script that is embedded within the container image. This is a shell script that can be used to run commands, tools or applications stored within the container image on container startup. We can inspect the container image’s run script using the singularity inspect command:

remote$ singularity inspect -r lolcow.sif

#!/bin/sh 

    date  |  cowsay  |  lolcat

This shows us the script within the lolcow.sif image configured to run by default when we use the singularity run command.

This seems very simple but already, we have downloaded a container image that is built with a different OS than is available on ARCHER2 that also contains software not available on ARCHER2 (cowsay and lolcat) and been able to run this on the ARCHER2 system without needing to install anything ourselves and without the container image having to know anything specific about how ARCHER2 is configured.

Key Points

  • Singularity is another container platform and it is often used in cluster/HPC/research environments.

  • Singularity has a different security model to other container platforms, one of the key reasons that it is well suited to HPC and cluster environments.

  • Singularity has its own container image format (SIF).

  • The singularity command can be used to pull images from Sylabs Cloud Library and run a container from an image file.

Using Singularity containers to run commands

Overview

Teaching: 10 min
Exercises: 5 min
Questions

  • How do I run different commands within a container?

  • How do I access an interactive shell within a container?

Objectives

  • Learn how to run different commands when starting a container.

  • Learn how to open an interactive shell within a container environment.

Running specific commands within a container

We saw earlier that we can use the singularity inspect command to see the run script that a container is configured to run by default. What if we want to run a different command within a container?

If we know the path of an executable that we want to run within a container, we can use the singularity exec command. For example, using the lolcow.sif container that we’ve already pulled from Singularity Hub, we can run the following within the test directory where the lolcow.sif file is located:

remote$ singularity exec lolcow.sif /bin/echo "Hello, world"

Hello, world

Here we see that a container has been started from the lolcow.sif image and the /bin/echo command has been run within the container, passing the input Hello, world. The command has echoed the provided input to the console and the container has terminated.

Note that the use of singularity exec has overriden any run script set within the image metadata and the command that we specified as an argument to singularity exec has been run instead.

Basic exercise: Running a different command within the “hello-world” container

Can you run a container based on the lolcow.sif image that prints the current date and time?

Solution

remote$ singularity exec lolcow.sif /bin/date

Fri Jun 26 15:17:44 BST 2020

Difference between singularity run and singularity exec

Above, we used the singularity exec command. In earlier episodes of this course we used singularity run. To clarify, the difference between these two commands is:

Opening an interactive shell within a container

If you want to open an interactive shell within a container, Singularity provides the singularity shell command. Again, using the lolcow.sif image, and within our test directory, we can run a shell within a container from the hello-world image:

remote$ singularity shell lolcow.sif

Singularity> whoami
[<your username>]
Singularity> ls
lolcow.sif
Singularity>

As shown above, we have opened a shell in a new container started from the lolcow.sif image. Note that the shell prompt has changed to show we are now within the Singularity container.

Use the exit command to exit from the container shell.

Key Points

  • The singularity exec is an alternative to singularity run that allows you to start a container running a specific command.

  • The singularity shell command can be used to start a container and run an interactive shell within it.

Using Docker images with Singularity

Overview

Teaching: 5 min
Exercises: 10 min
Questions

  • How do I use Docker images with Singularity?

Objectives

  • Learn how to run Singularity containers based on Docker images.

Using Docker images with Singularity

Singularity can also start containers directly from Docker container images, opening up access to a huge number of existing container images available on Docker Hub and other registries.

While Singularity doesn’t actually run a container using the Docker container image (it first converts it to a format suitable for use by Singularity), the approach used provides a seamless experience for the end user. When you direct Singularity to run a container based on a Docker container image, Singularity pulls the slices or layers that make up the Docker container image and converts them into a single-file Singularity SIF container image.

For example, moving on from the simple Hello World examples that we’ve looked at so far, let’s pull one of the official Docker Python container images. We’ll use the image with the tag 3.9.6-slim-buster which has Python 3.9.6 installed on Debian’s Buster (v10) Linux distribution:

remote$ singularity pull python-3.9.6.sif docker://python:3.9.6-slim-buster

INFO:    Converting OCI blobs to SIF format
INFO:    Starting build...
Getting image source signatures
Copying blob 33847f680f63 done  
Copying blob b693dfa28d38 done  
Copying blob ef8f1a8cefd1 done  
Copying blob 248d7d56b4a7 done  
Copying blob 478d2dfa1a8d done  
Copying config c7d70af7c3 done  
Writing manifest to image destination
Storing signatures
2021/07/27 17:23:38  info unpack layer: sha256:33847f680f63fb1b343a9fc782e267b5abdbdb50d65d4b9bd2a136291d67cf75
2021/07/27 17:23:40  info unpack layer: sha256:b693dfa28d38fd92288f84a9e7ffeba93eba5caff2c1b7d9fe3385b6dd972b5d
2021/07/27 17:23:40  info unpack layer: sha256:ef8f1a8cefd144b4ee4871a7d0d9e34f67c8c266f516c221e6d20bca001ce2a5
2021/07/27 17:23:40  info unpack layer: sha256:248d7d56b4a792ca7bdfe866fde773a9cf2028f973216160323684ceabb36451
2021/07/27 17:23:40  info unpack layer: sha256:478d2dfa1a8d7fc4d9957aca29ae4f4187bc2e5365400a842aaefce8b01c2658
INFO:    Creating SIF file...

Note how we see Singularity saying that it’s “Converting OCI blobs to SIF format”. We then see the layers of the Docker container image being downloaded and unpacked and written into a single SIF file. Once the process is complete, we should see the python-3.9.6.sif container image file in the current directory.

We can now run a container from this container image as we would with any other Singularity container image.

Running the Python 3.9.6 image that we just pulled from Docker Hub

Try running the Python 3.9.6 container image. What happens?

Try running some simple Python statements…

Running the Python 3.9.6 image

remote$ singularity run python-3.9.6.sif

This should put you straight into a Python interactive shell within the running container:

Python 3.9.6 (default, Jul 22 2021, 15:24:21) 
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>>

Now try running some simple Python statements:

>>> import math
>>> math.pi
3.141592653589793
>>>

In addition to running a container and having it run the default run script, you could also start a container running a shell in case you want to undertake any configuration prior to running Python. This is covered in the following exercise:

Open a shell within a Python container

Try to run a shell within a singularity container based on the python-3.9.6.sif container image. That is, run a container that opens a shell rather than the default Python interactive console as we saw above. See if you can find more than one way to achieve this.

Within the shell, try starting the Python interactive console and running some Python commands.

Solution

Recall from the earlier material that we can use the singularity shell command to open a shell within a container. To open a regular shell within a container based on the python-3.9.6.sif container image, we can therefore simply run:

remote$ singularity shell python-3.9.6.sif

Singularity> echo $SHELL
/bin/bash
Singularity> cat /etc/issue
Debian GNU/Linux 10 \n \l

Singularity> python
Python 3.9.6 (default, Jul 22 2021, 15:24:21) 
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> print('Hello World!')
Hello World!
>>> exit()

Singularity> exit
$

It is also possible to use the singularity exec command to run an executable within a container. We could, therefore, use the exec command to run /bin/bash:

remote$ singularity exec python-3.9.6.sif /bin/bash

Singularity> echo $SHELL
/bin/bash

You can run the Python console from your container shell simply by running the python command.

References

[1] Gregory M. Kurzer, Containers for Science, Reproducibility and Mobility: Singularity P2. Intel HPC Developer Conference, 2017

Key Points

  • Singularity can start a container from a Docker image which can be pulled directly from Docker Hub.

Break

Overview

Teaching: min
Exercises: min
Questions
Objectives

Lunch break

Key Points

The Singularity cache

Overview

Teaching: 10 min
Exercises: 0 min
Questions

  • Why does Singularity use a local cache?

  • Where does Singularity store images?

Objectives

  • Learn about Singularity’s image cache.

  • Learn how to manage Singularity images stored locally.

Singularity’s image cache

Singularity uses a local cache to save downloaded container image files in addition to storing them as the file you specify. As we saw in the previous episode, images are simply .sif files stored on your local disk.

If you delete a local .sif container image that you have pulled from a remote container image repository and then pull it again, if the container image is unchanged from the version you previously pulled, you will be given a copy of the container image file from your local cache rather than the container image being downloaded again from the remote source. This removes unnecessary network transfers and is particularly useful for large container images which may take some time to transfer over the network. To demonstrate this, remove the lolcow.sif file stored in your test directory and then issue the pull command again:

remote$ rm lolcow.sif
remote$ singularity pull lolcow.sif library://lolcow

INFO:    Using cached image

As we can see in the above output, the container image has been returned from the cache and we do not see the output that we saw previously showing the container image being downloaded from the Cloud Library.

How do we know what is stored in the local cache? We can find out using the singularity cache command:

remote$ singularity cache list

There are 2 container file(s) using 129.35 MiB and 7 oci blob file(s) using 41.49 MiB of space
Total space used: 170.84 MiB

This tells us how many container image files are stored in the cache and how much disk space the cache is using but it doesn’t tell us what is actually being stored. To find out more information we can add the -v verbose flag to the list command:

remote$ singularity cache list -v

There are 2 container file(s) using 129.35 MiB and 7 oci blob file(s) using 41.49 MiB of space
Total space used: 170.84 MiB
artta118@ln04:~/test> singularity cache list -v
NAME                     DATE CREATED           SIZE             TYPE
50b2668d8d3f74c49a7280   2023-09-12 11:41:31    0.96 KiB         blob
76f124aca9afaf3f75812d   2023-09-12 11:41:30    2.51 MiB         blob
7becefa709e2358336177a   2023-09-12 11:41:31    6.25 KiB         blob
87bc5aa6fc4253b93dee0a   2023-09-12 11:41:30    0.23 KiB         blob
dae1d9fd74c12f7e66b92c   2023-09-12 11:41:29    10.43 MiB        blob
e1acddbe380c63f0de4b77   2023-09-12 11:41:27    25.89 MiB        blob
ecc7ff4d26223f4545c4fd   2023-09-12 11:41:28    2.64 MiB         blob
sha256.cef378b9a9274c2   2023-09-12 11:39:18    90.43 MiB        library
28bed4c51c3b531159d8af   2023-09-12 11:41:36    38.92 MiB        oci-tmp

There are 2 container file(s) using 129.35 MiB and 7 oci blob file(s) using 41.49 MiB of space
Total space used: 170.84 MiB

This provides us with some more useful information about the actual container images stored in the cache. In the TYPE column we can see that our container image type is library because it’s a SIF container image that has been pulled from the Cloud Library.

Cleaning the Singularity image cache

We can remove container images from the cache using the singularity cache clean command. Running the command without any options will display a warning and ask you to confirm that you want to remove everything from your cache.

You can also remove specific container images or all container images of a particular type. Look at the output of singularity cache clean --help for more information.

Cache location

By default, Singularity uses $HOME/.singularity/cache as the location for the cache. You can change the location of the cache by setting the SINGULARITY_CACHEDIR environment variable to the cache location you want to use.

Key Points

  • Singularity caches downloaded images so that an unchanged image isn’t downloaded again when it is requested using the singularity pull command.

  • You can free up space in the cache by removing all locally cached images or by specifying individual images to remove.

Files in Singularity containers

Overview

Teaching: 10 min
Exercises: 10 min
Questions

  • How do I make data available in a Singularity container?

  • What data is made available by default in a Singularity container?

Objectives

  • Understand that some data from the host system is usually made available by default within a container

  • Learn more about how Singularity handles users and binds directories from the host filesystem.

The key concept to remember when running a Singularity container, you only have the same permissions to access files as the user on the host system that you start the container as. (If you are familiar with Docker, you may note that this is different behaviour than you would see with that tool.)

In this episode we will look at working with files in the context of Singularity containers and how this links with Singularity’s approach to users and permissions within containers.

Users within a Singularity container

The first thing to note is that when you ran whoami within the container shell you started at the end of the previous episode, you should have seen the same username that you have on the host system when you ran the container.

For example, if my username were jc1000, I would expect to see the following:

remote$ singularity shell lolcow.sif
Singularity> whoami

artta118

But wait! I downloaded the standard, public version of the lolcow container image from the Cloud Library. I haven’t customised it in any way. How is it configured with my own user details?!

If you have any familiarity with Linux system administration, you may be aware that in Linux, users and their Unix groups are configured in the /etc/passwd and /etc/group files respectively. In order for the running container to know of my user, the relevant user information needs to be available within these files within the container.

Assuming this feature is enabled within the installation of Singularity on your system, when the container is started, Singularity appends the relevant user and group lines from the host system to the /etc/passwd and /etc/group files within the container[1].

This means that the host system can effectively ensure that you cannot access/modify/delete any data you should not be able to on the host system from within the container and you cannot run anything that you would not have permission to run on the host system since you are restricted to the same user permissions within the container as you are on the host system.

Files and directories within a Singularity container

Singularity also binds some directories from the host system where you are running the singularity command into the container that you are starting. Note that this bind process is not copying files into the running container, it is making an existing directory on the host system visible and accessible within the container environment. If you write files to this directory within the running container, when the container shuts down, those changes will persist in the relevant location on the host system.

There is a default configuration of which files and directories are bound into the container but ultimate control of how things are set up on the system where you are running Singularity is determined by the system administrator. As a result, this section provides an overview but you may find that things are a little different on the system that you’re running on.

One directory that is likely to be accessible within a container that you start is your home directory. You may also find that the directory from which you issued the singularity command (the current working directory) is also bound.

The binding of file content and directories from a host system into a Singularity container is illustrated in the example below showing a subset of the directories on the host Linux system and in a running Singularity container:

Host system:                                                      Singularity container:
-------------                                                     ----------------------
/                                                                 /
├── bin                                                           ├── bin
├── etc                                                           ├── etc
│   ├── ...                                                       │   ├── ...
│   ├── group  ─> user's group added to group file in container ─>│   ├── group
│   └── passwd ──> user info added to passwd file in container ──>│   └── passwd
├── home                                                          ├── usr
│   └── artta118 ───> user home directory made available ──> ─┐   ├── sbin
├── usr                 in container via bind mount         │     ├── home
├── sbin                                                    └────────>└── artta118
└── ...                                                           └── ...

Questions and exercises: Files in Singularity containers

Q1: What do you notice about the ownership of files in a container started from the lolcow.sif image? (e.g. take a look at the ownership of files in the root directory (/) and your home directory (~/)).

Exercise 1: In this container, try creating a file in the root directory / (e.g. using touch /myfile.dat). What do you notice? Try removing the /singularity file. What happens in these two cases?

Exercise 2: In your home directory within the container shell, try and create a simple text file (e.g. echo "Some text" > ~/test-file.txt). Is it possible to do this? If so, why? If not, why not?! If you can successfully create a file, what happens to it when you exit the shell and the container shuts down?

Answers

A1: Use the ls -l / command to see a detailed file listing including file ownership and permission details. You should see that most of the files in the / directory are owned by root, as you would probably expect on any Linux system. If you look at the files in your home directory, they should be owned by you.

A Ex1: We’ve already seen from the previous answer that the files in / are owned by root so we would nott expect to be able to create files there if we’re not the root user. However, if you tried to remove /singularity you would have seen an error similar to the following: cannot remove '/singularity': Read-only file system. This tells us something else about the filesystem. It’s not just that we do not have permission to delete the file, the filesystem itself is read-only so even the root user would not be able to edit/delete this file. We will look at this in more detail shortly.

A Ex2: Within your home directory, you should be able to successfully create a file. Since you’re seeing your home directory on the host system which has been bound into the container, when you exit and the container shuts down, the file that you created within the container should still be present when you look at your home directory on the host system.

Binding additional host system directories to the container

You will sometimes need to bind additional host system directories into a container you are using over and above those bound by default. For example:

The -B option to the singularity command is used to specify additional binds. For example, to bind the /opt/cray directory (where the HPE Cray programming environment is stored) into a container you could use:

remote$ singularity shell -B /opt/cray lolcow.sif
Singularity> ls -la /opt/cray

Note that, by default, a bind is mounted at the same path in the container as on the host system. You can also specify where a host directory is mounted in the container by separating the host path from the container path by a colon (:) in the option:

remote$ singularity shell -B /opt/cray:/cpe lolcow.sif
Singularity> ls -la /cpe

You can specify multiple binds to -B by separating them by commas (,).

Another option is to specify the paths you want to bind in the SINGULARITY_BIND environment variable. This can be more convenient when you have a lot of paths you want to bind into the running container (we will see this later in the course when we look at using MPI with containers). For example, to bind the locations that contain both the HPE Cray programming environment and the CSE centrally installed software into a running container, we would use:

remote$ export SINGULARITY_BIND="/opt/cray,/work/y07/shared"
remote$ singularity shell lolcow.sif
Singularity> ls -la /work/y07/shared

Finally, you can also copy data into a container image at build time if there is some static data required in the image. We cover this later in the section on building container images.

References

[1] Gregory M. Kurzer, Containers for Science, Reproducibility and Mobility: Singularity P2. Intel HPC Developer Conference, 2017. Available at: https://www.intel.com/content/dam/www/public/us/en/documents/presentation/hpc-containers-singularity-advanced.pdf

Key Points

  • Your current directory and home directory are usually available by default in a container.

  • You have the same username and permissions in a container as on the host system.

  • You can specify additional host system directories to be available in the container.

Creating Your Own Container Images

Overview

Teaching: 20 min
Exercises: 15 min
Questions

  • How can I make my own Docker container images?

  • How do I document the ‘recipe’ for a Docker container image?

Objectives

  • Explain the purpose of a Dockerfile and show some simple examples.

  • Demonstrate how to build a Docker container image from a Dockerfile.

  • Compare the steps of creating a container image interactively versus a Dockerfile.

  • Create an installation strategy for a container image.

  • Demonstrate how to upload (‘push’) your container images to the Docker Hub.

  • Describe the significance of the Docker Hub naming scheme.

There are lots of reasons why you might want to create your own container images.

Why Docker, not Singularity?

In this section we will use Docker, rather than Singularity, to create our own container images. However, we will still use Singularity on the remote HPC platform to run containers based on the container images we have created. Why do we use a separate tool to create our container images rather than just using Singularity to build container images too?

One of the most important points to note is that, generally you must have administrator privileges on the system where you want to build container images. On the remote HPC system where we have been running Singularity containers we do not have administrator privileges so we cannot use these systems to build our own container images. In this course, we will use our laptop/workstation to build container images though it is also possible ot build them on commercial cloud resources and through other systems such as Github actions - using these more advanced options is beyond the scope of this course.

This still does not explain why we are using Docker rather than Singularity to build our container images. Using Docker on our laptop/workstation has a number of advantages:

As we have already seen, Docker images can be used by Singularity (it converts them to Singularity image format automatically) so using Docker to build container images is typically an easier option.

More on Docker?

You can find a full introduction to Docker in the Introduction to Docker lesson in the Carpentries Incubator. Here, we will use a subset of that lesson to build our own container images.

Interactive installation

We could, potentially, install software in a running container and then save the updated state of the running container as a new container image but we are not going to use this approach to create our own container images. The main reason why we do not do this is that it it reduces the reproducibility of our images. It is also not typical practice for creating container images.

Instead of installing software interactively, we are going to create our container image from a reproducible recipe, known as a Dockerfile.

Check that Docker is working

As mentioned above, we will be building container images on your local system (usually your laptop) rather than on the remote HPC system. You can keep your terminal to the remote system open as we will be using later to run containers based on the images we are creating. Open a new terminal on your laptop.

Start the Docker application that you installed in working through the setup instructions for the workshop. Note that this might not be necessary if your laptop is running Linux or if the installation added the Docker application to your startup process.

You may need to login to Docker Hub

The Docker application will usually provide a way for you to log in to the Docker Hub using the application’s menu (macOS) or systray icon (Windows) and it is usually convenient to do this when the application starts. This will require you to use your Docker Hub username and your password. We will not actually require access to the Docker Hub until later in the course but if you can login now, you should do so.

Determining your Docker Hub username

If you no longer recall your Docker Hub username, e.g., because you have been logging into the Docker Hub using your email address, you can find out what it is through the steps:

  • Open https://hub.docker.com/ in a web browser window
  • Sign-in using your email and password (don’t tell us what it is)
  • In the top-right of the screen you will see your username

Once your Docker application is running, open a shell (terminal) window, and run the following command to check that Docker is installed and the command line tools are working correctly. Below is the output for a Mac version, but the specific version is unlikely to matter much: it does not have to precisely match the one listed below.

$ docker --version

Docker version 20.10.5, build 55c4c88

The above command has not actually relied on the part of Docker that runs containers, just that Docker is installed and you can access it correctly from the command line.

A command that checks that Docker is working correctly is the docker container ls command (we cover this command in more detail later in the course).

Without explaining the details, output on a newly installed system would likely be:

$ docker container ls

CONTAINER ID        IMAGE               COMMAND             CREATED             STATUS              PORTS               NAMES

(The command docker system info could also be used to verify that Docker is correctly installed and operational but it produces a larger amount of output.)

However, if you instead get a message similar to the following

Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?

then you need to check that you have started the Docker Desktop, Docker Engine, or however else you worked through the setup instructions.

Put installation instructions in a Dockerfile

As mentioned above, building images using Docker should be done by adding the build and installation instructions into a Dockerfile.

A Dockerfile is a plain text file with keywords and commands that can be used by Docker to create a new container image. In this first example, we are going to create a container image based on the lightweight Alpine Linux distribution with Python 3 added.

From your shell, create a new folder to contain the Dockerfile`. Typically, all the data required to build a container image is placed in the same folder - you will have one folder per container image you build.

$ mkdir alpine-python
$ cd alpine-python

Once you are in the directory, use your favourite text editor to create the Dockerfile to build the new container image. You should create a file called Dockerfile with the following contents.

FROM alpine
RUN apk add --update python3 py3-pip python3-dev
RUN pip install cython
CMD ["python3", "--version"]

Let’s break this file down:

shell-form and exec-form for CMD

Another way to specify the parameter for the CMD instruction is the shell-form. Here you type the command as you would call it from the command line. Docker then silently runs this command in the image’s standard shell. CMD cat /etc/passwd is equivalent to CMD ["/bin/sh", "-c", "cat /etc/passwd"]. We recommend to prefer the more explicit exec-form because we will be able to create more flexible container image command options and make sure complex commands are unambiguous in this format.

Create a new Docker container image

So far, we only have a text file named Dockerfile – we do not yet have a container image. We want Docker to take this Dockerfile, run the installation commands contained within it, and then save the resulting container as a new container image. To do this we will use the docker image build command.

We have to provide docker image build with three pieces of information:

All together, the build command that you should run on your computer, will have a similar structure to this:

$ docker image build --platform linux/amd64 -t USERNAME/CONTAINER_IMAGE_NAME .

The -t option names the container image; the final dot indicates that the Dockerfile is in our current directory.

For example, if my user name was alice and I wanted to call my container image alpine-python, I would use this command:

$ docker image build --platform linux/amd64 -t alice/alpine-python .

## Build your container image

Based on the information above, try using Docker to build your container image.

## Solution

Using a command like:

$ docker image build --platform linux/amd64 -t alice/alpine-python .

should build the new container image. while building, you will see output similar to:

[+] Building 21.4s (7/7) FINISHED                                                                           
docker:desktop-linux
=> [internal] load .dockerignore  0.0s
=>
=> transferring context: 2B  0.0s
=> [internal] load build definition from Dockerfile  0.0s
=>
=> transferring dockerfile: 159B  0.0s
=> [internal] load metadata for docker.io/library/alpine:latest 1.3s
=> [1/3] FROM docker.io/library/alpine:latest@sha256:7144f7bab3d4c2648d7e59409f15ec52a18006a128c733fcff20d3a4a54ba44a 0.4s
=>
=> resolve docker.io/library/alpine:latest@sha256:7144f7bab3d4c2648d7e59409f15ec52a18006a128c733fcff20d3a4a54ba44a 0.0s
=>
=> sha256:7144f7bab3d4c2648d7e59409f15ec52a18006a128c733fcff20d3a4a54ba44a 1.64kB / 1.64kB 0.0s
=>
=> sha256:c5c5fda71656f28e49ac9c5416b3643eaa6a108a8093151d6d1afc9463be8e33 528B / 528B 0.0s
=>
=> sha256:7e01a0d0a1dcd9e539f8e9bbd80106d59efbdf97293b3d38f5d7a34501526cdb 1.47kB / 1.47kB 0.0s
=>
=> sha256:7264a8db6415046d36d16ba98b79778e18accee6ffa71850405994cffa9be7de 3.40MB / 3.40MB 0.3s
=>
=> extracting sha256:7264a8db6415046d36d16ba98b79778e18accee6ffa71850405994cffa9be7de     0.1s
=> [2/3] RUN apk add --update python3 py3-pip python3-dev                                 8.2s
=> [3/3] RUN pip install cython                                                           0.3s 
=>
=> exporting layers                                                                       0.3s 
=>
=> writing image sha256:b317a6d5b97daee5055cc4cb4432dd6cfe98a9e6f67ed5b3036fda608c629d23  0.0s 
=>
=> naming to docker.io/atunerepcc/alpine-python                                           0.0s 

What's Next?                          
  View summary of image vulnerabilities and recommendations → docker scout quickview

Build Context

Notice that the final input to docker image build isn’t the Dockerfile – it’s a directory! In the command above, we’ve used the current working directory (.) of the shell as the final input to the docker image build command. This option provides what is called the build context to Docker – if there are files being copied into the built container image more details in the next episode they’re assumed to be in this location. Docker expects to see a Dockerfile in the build context also (unless you tell it to look elsewhere).

Even if it will not need all of the files in the build context directory, Docker does “load” them before starting to build, which means that it’s a good idea to have only what you need for the container image in a build context directory, as we’ve done in this example.

Listing the images you have locally on your laptop

To show the images you have available in Docker on your laptop (hopefully including the one you have just built!) you use the docker image ls command:

$ docker image ls

REPOSITORY                   TAG               IMAGE ID       CREATED          SIZE
alice/alpine-python          latest            aa0400bf60c3   14 seconds ago   166MB

Share your new container image on Docker Hub

Container images that you release publicly can be stored on the Docker Hub for free. If you name your container image as described above, with your Docker Hub username, all you need to do is run the opposite of docker image pull – docker image push.

$ docker image push alice/alpine-python

Make sure to substitute the full name of your container image!

In a web browser, open https://hub.docker.com, and on your user page you should now see your container image listed, for anyone to use or build on.

Logging In

Technically, you have to be logged into Docker on your computer for this to work. Usually it happens by default, but if docker image push doesn’t work for you, run docker login first, enter your Docker Hub username and password, and then try docker image push again.

Exercise: Run your new container image using Singularity

Now your container image is on Docker Hub, you should be able to use what you learned earlier to use Singularity download the new container image to the HPC platform, convert it to a Singularity SIF file and run a container based on the image.

  1. What command (on the remote HPC system) will download your new container image from Docker Hub and save to a Singularity image file (SIF)?
  2. What command would you then use to run a container based on your new container image on the remote HPC system?
  3. Can you make your running container do something different, like print “hello world”?

Solution

  1. To download your new image from Docker Hub and save it as a SIF file called alpine-python.sif you would use something like

    remote$ singularity pull alpine-python.sif docker://alice/alpine-python

    (remember to use your Docker Hub username instead of alice!).

  2. To run the default command using the new container image, you would use

    remote$ singularity run alpine-python.sif

    This should print the version of Python in the container.

  3. To run a different command, we would use something like:

    remote$ singluarity exec alpine-python.sif echo "hello world"

While it may not look like you have achieved much, you have already effected the combination of a lightweight Linux operating system with your specification to run a given command that can operate reliably on macOS, Microsoft Windows, Linux and on a remote HPC platform!

Boring but important notes about installation

There are a lot of choices when it comes to installing software – sometimes too many! Here are some things to consider when creating your own container image:

In general, a good strategy for installing software is:

What’s in a name? (again)

You don’t have to name your containers images using the USERNAME/CONTAINER_IMAGE_NAME:TAG naming scheme. On your own computer, you can call container images whatever you want, and refer to them by the names you choose. It’s only when you want to share a container image that it needs the correct naming format.

You can rename container images using the docker image tag command. For example, imagine someone named Alice has been working on a workflow container image and called it workflow-test on her own computer. She now wants to share it in her alice Docker Hub account with the name workflow-complete and a tag of v1. Her docker image tag command would look like this:

$ docker image tag workflow-test alice/workflow-complete:v1

She could then push the re-named container image to Docker Hub, using docker image push alice/workflow-complete:v1

Key Points

  • Dockerfiles specify what is within Docker container images.

  • The docker image build command is used to build a container image from a Dockerfile.

  • You can share your Docker container images through the Docker Hub so that others can create Docker containers from your container images.

Lunch

Overview

Teaching: min
Exercises: min
Questions
Objectives

Lunch break

Key Points

Creating More Complex Container Images

Overview

Teaching: 15 min
Exercises: 20 min
Questions

  • How can I make more complex container images?

Objectives

  • Explain how you can include files within Docker container images when you build them.

  • Explain how you can access files on the Docker host from your Docker containers.

Building container images with your files included

In order to create and use your own container images, you may need more information than our previous example. You may want to use files from outside the container, that are not included within the container image by copying the files into the container image at build time.

Create a Python script

Before we go ahead and build our next container image, we’re going to create a simple Python script on our host system and create a Dockerfile to have this script copied into our container image when it is created.

In your shell, create a new directory to hold the build context for our new container image and move into the directory:

$ mkdir alpine-sum
$ cd alpine-sum

Use your text editor to create a Python script called sum.py with the following contents:

#!/usr/bin/env python3

import sys
try:
   total = sum(int(arg) for arg in sys.argv[1:])
   print('sum =', total)
except ValueError:
   print('Please supply integer arguments')

Let’s assume that we’ve finished with our sum.py script and want to add it to the container image itself.

Create the Dockerfile

Now we have our Python script, we are going to create our Dockerfile. This is going to be similar to the Dockerfile we used in the previous section with the addition of one extra line. Here is the full Dockerfile:

FROM alpine
RUN apk add --update python3 py3-pip python3-dev
RUN pip install cython
COPY sum.py /home
CMD ["python3", "--version"]

The additional line we have added is:

COPY sum.py /home

This line will cause Docker to copy the file from your computer into the container’s filesystem. Let’s build the container image like before, but give it a different name and then push it to Docker Hub (remember to subsitute alice for your Docker Hub username):

$ docker image build --platform linux/amd64 -t alice/alpine-sum .

...output from docker build...

$ docker push alice/alpine-sum

The Importance of Command Order in a Dockerfile

When you run docker build it executes the build in the order specified in the Dockerfile. This order is important for rebuilding and you typically will want to put your RUN commands before your COPY commands.

Docker builds the layers of commands in order. This becomes important when you need to rebuild container images. If you change layers later in the Dockerfile and rebuild the container image, Docker doesn’t need to rebuild the earlier layers but will instead used a stored (called “cached”) version of those layers.

For example, in an instance where you wanted to copy multiply.py into the container image instead of sum.py. If the COPY line came before the RUN line, it would need to rebuild the whole image. If the COPY line came second then it would use the cached RUN layer from the previous build and then only rebuild the COPY layer.

Exercise: Did it work?

Can you remember how to run a container interactively on the remote HPC system? Try that with this one. Once inside, try running the Python script you added to the container image.

Solution

You can start the container interactively on the remote HPC system like so (remember to use your Docker Hub username):

remote$ singularity pull alpine-sum.sif docker://alice/alpine-sum
remote$ singularity shell alpine-sum.sif

You should be able to run the python command inside the container like this:

Singularity> python3 /home/sum.py

This COPY keyword can be used to place your own scripts or own data into a container image that you want to publish or use as a record. Note that it’s not necessarily a good idea to put your scripts inside the container image if you’re constantly changing or editing them. Then, referencing the scripts from outside the container is a good idea, by bind mounting host data into the running container as we saw earlier in the workshop. You also want to think carefully about size – if you run docker image ls you’ll see the size of each container image all the way on the right of the screen. The bigger your container image becomes, the harder it will be to easily download.

Security Warning

Login credentials including passwords, tokens, secure access tokens or other secrets must never be stored in a container. If secrets are stored, they are at high risk to be found and exploited when made public.

Copying alternatives

Another trick for getting your own files into a container image is by using the RUN keyword and downloading the files from the internet. For example, if your code is in a GitHub repository, you could include this statement in your Dockerfile to download the latest version every time you build the container image:

RUN git clone https://github.com/alice/mycode

Similarly, the wget command can be used to download any file publicly available on the internet:

RUN wget ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/2.10.0/ncbi-blast-2.10.0+-x64-linux.tar.gz

Note that the above RUN examples depend on commands (git and wget respectively) that must be available within your container: Linux distributions such as Alpine may require you to install such commands before using them within RUN statements.

More fancy Dockerfile options (optional, for presentation or as exercises)

We can expand on the example above to make our container image even more “automatic”. Here are some ideas:

Make the sum.py script run automatically

FROM alpine
RUN apk add --update python3 py3-pip python3-dev
COPY sum.py /home

# Run the sum.py script as the default command
CMD ["python3", "/home/sum.py"]

Build and push this:

$ docker image build --platform linux/amd64 -t alice/alpine-sum:v1 .
$ docker push alice/alpine-sum:v1

You’ll notice that you can run the container without arguments just fine, resulting in sum = 0, but this is boring. Supplying arguments however doesn’t work:

remote$ singularity pull alpine-sum_v1.sif docker://alice/alpine-sum:v1
remote$ singularity run alpine-sum_v1.sif 10 11 12

results in:

FATAL:   "10": executable file not found in $PATH

This is because the arguments 10 11 12 are ignored by the CMD syntax and the container tries to interpret them as additional commands instead of arguments to the sum.py script.

To achieve the goal of having a command that always runs when a container is run from the container image and can be passed the arguments given on the command line, use the keyword ENTRYPOINT in the Dockerfile.

FROM alpine

COPY sum.py /home
RUN apk add --update python3 py3-pip python3-dev

# Run the sum.py script as the default command and
# allow people to enter arguments for it
ENTRYPOINT ["python3", "/home/sum.py"]

# Give default arguments, in case none are supplied on
# the command-line
CMD ["10", "11"]

Build and push this:

$ docker image build --platform linux/amd64 -t alice/alpine-sum:v2 .
$ docker push alice/alpine-sum:v2

remote$ singularity pull alpine-sum_v2.sif docker://alice/alpine-sum:v2
remote$ singularity run alpine-sum_v2.sif 10 11 12

sum = 33

Add the sum.py script to the PATH so you can run it directly:

FROM alpine

RUN apk add --update python3 py3-pip python3-dev

COPY sum.py /home
# set script permissions
RUN chmod +x /home/sum.py
# add /home folder to the PATH
ENV PATH /home:$PATH

Build and push this:

$ docker image build --platform linux/amd64 -t alice/alpine-sum:v3 .
$ docker push alice/alpine-sum:v3

Build and test it:

remote$ singularity pull alpine-sum_v3.sif docker://alice/alpine-sum:v3
remote$ singularity exec alpine-sum_v3.sif sum.py 10 11 12

sum = 33

Best practices for writing Dockerfiles

Take a look at Nüst et al.’s “Ten simple rules for writing Dockerfiles for reproducible data science” [1] for some great examples of best practices to use when writing Dockerfiles. The GitHub repository associated with the paper also has a set of example Dockerfiles demonstrating how the rules highlighted by the paper can be applied.

[1] Nüst D, Sochat V, Marwick B, Eglen SJ, Head T, et al. (2020) Ten simple rules for writing Dockerfiles for reproducible data science. PLOS Computational Biology 16(11): e1008316. https://doi.org/10.1371/journal.pcbi.1008316

Key Points

  • Docker allows containers to read and write files from the Docker host.

  • You can include files from your Docker host into your Docker container images by using the COPY instruction in your Dockerfile.

Running MPI parallel jobs using Singularity containers

Overview

Teaching: 30 min
Exercises: 40 min
Questions

  • How do I set up and run an MPI job from a Singularity container?

Objectives

  • Learn how MPI applications within Singularity containers can be run on HPC platforms

  • Understand the challenges and related performance implications when running MPI jobs via Singularity

What is MPI?

MPI - Message Passing Interface - is a widely used standard for parallel programming. It is used for exchanging messages/data between processes in a parallel application. If you’ve been involved in developing or working with computational science software.

Usually, when working on HPC systems, you compile your application against the MPI libraries provided on the system or you use applications that have been compiled by the HPC system support team. This approach to portability: source code portability is the traditional approach to making applications portable to different HPC platforms.

However, compiling complex HPC applications that have lots of dependencies (including MPI) is not always straightforward and can be a significant challenge as most HPC systems differ in various ways in terms of OS and base software available. There are a number of different approaches that can be taken to make it easier to deploy applications on HPC systems; for example, the Spack software automates the dependency resolution and compilation of applications. Containers provide another potential way to resolve these problems but care needs to be taken when interfacing with MPI on the host system which adds more complexity to running containers in parallel on HPC systems.

MPI codes with Singularity containers

Obviously, we will not have admin/root access on the HPC platform we are using so cannot (usually) build our container images on the HPC system itself. However, we do need to ensure our container is using the MPI library on the HPC system itself so we can get the performance benefit of the HPC interconnect. How do we overcome these contradictions?

The answer is that we install a version of the MPI library in our container image that is binary compatible with the MPI library on the host system and install our software in the container image using the local version of the MPI library. At runtime, we then ensure that the MPI library from the host is used within the running container rather than the locally-installed version of MPI.

There are two widely used open source MPI library distributions on HPC systems:

This typically means that if you want to distribute HPC software that uses MPI within a container image you will need to maintain versions that are compatible with both MPICH and OpenMPI. There are efforts underway to provide tools that will provide a binary interface between different MPI implementations, e.g. HPE Cray’s MPIxlate software but these are not generally available yet.

Building a Singularity container image with MPI software

This example makes the assumption that you’ll be building a container image on a local platform and then deploying it to a HPC system with a different but compatible MPI implementation using a combination of the Hybrid and Bind models from the Singularity documentation. We will build our application using MPI in the container image but will bind the MPI library from the host into the container at runtime. See Singularity and MPI applications in the Singularity documentation for more technical details.

The example we will build will:

The Dockerfile to install MPICH and the OSU micro-benchmark we will use to build the container image is shown below. Create a new directory called osu-benchmarks to hold the build context for this new image. Create the Dockerfile in this directory.

FROM ubuntu:20.04

ENV DEBIAN_FRONTEND=noninteractive

# Install the necessary packages (from repo)
RUN apt-get update && apt-get install -y --no-install-recommends \
 apt-utils \
 build-essential \
 curl \
 libcurl4-openssl-dev \
 libzmq3-dev \
 pkg-config \
 software-properties-common
RUN apt-get clean
RUN apt-get install -y dkms
RUN apt-get install -y autoconf automake build-essential numactl libnuma-dev autoconf automake gcc g++ git libtool

# Download and build an ABI compatible MPICH
RUN curl -sSLO http://www.mpich.org/static/downloads/3.4.2/mpich-3.4.2.tar.gz \
   && tar -xzf mpich-3.4.2.tar.gz -C /root \
   && cd /root/mpich-3.4.2 \
   && ./configure --prefix=/usr --with-device=ch4:ofi --disable-fortran \
   && make -j8 install \
   && cd / \
   && rm -rf /root/mpich-3.4.2 \
   && rm /mpich-3.4.2.tar.gz

# OSU benchmarks
RUN curl -sSLO http://mvapich.cse.ohio-state.edu/download/mvapich/osu-micro-benchmarks-5.4.1.tar.gz \
   && tar -xzf osu-micro-benchmarks-5.4.1.tar.gz -C /root \
   && cd /root/osu-micro-benchmarks-5.4.1 \
   && ./configure --prefix=/usr/local CC=/usr/bin/mpicc CXX=/usr/bin/mpicxx \
   && cd mpi \
   && make -j8 install \
   && cd / \
   && rm -rf /root/osu-micro-benchmarks-5.4.1 \
   && rm /osu-micro-benchmarks-5.4.1.tar.gz

# Add the OSU benchmark executables to the PATH
ENV PATH=/usr/local/libexec/osu-micro-benchmarks/mpi/startup:$PATH
ENV PATH=/usr/local/libexec/osu-micro-benchmarks/mpi/pt2pt:$PATH
ENV PATH=/usr/local/libexec/osu-micro-benchmarks/mpi/collective:$PATH
ENV OSU_DIR=/usr/local/libexec/osu-micro-benchmarks/mpi

# path to mlx IB libraries in Ubuntu
ENV LD_LIBRARY_PATH=/usr/lib/libibverbs:$LD_LIBRARY_PATH

A quick overview of what the above definition file is doing:

Build and test the OSU Micro-Benchmarks image

Using the above Dockerfile, build a container image and push it to Docker Hub.

Pull the image on ARCHER2 using Singularity to convert it to a Singularity image and the test it by running the osu_hello benchmark that is found in the startup benchmark folder with either singularity exec or singularity shell.

Note: the build process can take a while. If you want to test running while the build is happening, you can log into ARCHER2 and use a pre-built version of the container image to test. You can find this container image at:

${EPCC_SINGULARITY_DIR}/osu_benchmarks.sif

Solution

You should be able to build an image and push it to Docker Hub as follows:

$ docker image build --platform linux/amd64 -t alice/osu-benchmarks
$ docker push alice/osu-benchmarks

You can then log into ARCHER2 and pull the container image from Docker Hub with:

remote$ singularity pull osu-benchmarks.sif docker://alice/osu-benchmarks

Let’s begin with a single-process run of startup/osu_hello to ensure that we can run the container as expected. We’ll use the MPI installation within the container for this test. Note that when we run a parallel job on an HPC cluster platform, we use the MPI installation on the cluster to coordinate the run so things are a little different… we will see this shortly.

Start a shell in the Singularity container based on your image and then run a single process job via mpirun:

$ singularity shell --contain osu_benchmarks.sif
Singularity> mpirun -np 1 osu_hello

You should see output similar to the following:

# OSU MPI Hello World Test v5.7.1
This is a test with 1 processes

Running Singularity containers with MPI on HPC system

Assuming the above tests worked, we can now try undertaking a parallel run of one of the OSU benchmarking tools within our container image on the remote HPC platform.

This is where things get interesting and we will begin by looking at how Singularity containers are run within an MPI environment.

If you’re familiar with running MPI codes, you’ll know that you use mpirun (as we did in the previous example), mpiexec, srun or a similar MPI executable to start your application. This executable may be run directly on the local system or cluster platform that you’re using, or you may need to run it through a job script submitted to a job scheduler. Your MPI-based application code, which will be linked against the MPI libraries, will make MPI API calls into these MPI libraries which in turn talk to the MPI daemon process running on the host system. This daemon process handles the communication between MPI processes, including talking to the daemons on other nodes to exchange information between processes running on different machines, as necessary.

When running code within a Singularity container, we don’t use the MPI executables stored within the container, i.e. we DO NOT run:

singularity exec mpirun -np <numprocs> /path/to/my/executable

Instead we use the MPI installation on the host system to run Singularity and start an instance of our executable from within a container for each MPI process. Without Singularity support in an MPI implementation, this results in starting a separate Singularity container instance within each process. This can present some overhead if a large number of processes are being run on a host. Where Singularity support is built into an MPI implementation this can address this potential issue and reduce the overhead of running code from within a container as part of an MPI job.

Ultimately, this means that our running MPI code is linking to the MPI libraries from the MPI install within our container and these are, in turn, communicating with the MPI daemon on the host system which is part of the host system’s MPI installation. In the case of MPICH, these two installations of MPI may be different but as long as there is ABI compatibility between the version of MPI installed in your container image and the version on the host system, your job should run successfully.

We can now try running a 2-process MPI run of a point to point benchmark osu_latency on ARCHER2.

Undertake a parallel run of the osu_latency benchmark (general example)

Create a job submission script called submit.slurm on the /work file system on ARCHER2 to run containers based on the container image across two nodes on ARCHER2. The example below uses the osu-benchmarks container image that is already available on ARCHER2 but you can easily modify it to use your version of the container image if you wish - the results should be the same in both cases.

A template based on the example in the ARCHER2 documentation is:

#!/bin/bash

#SBATCH --job-name=singularity_parallel
#SBATCH --time=0:10:0
#SBATCH --nodes=2
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=1

#SBATCH --partition=standard
#SBATCH --qos=short
#SBATCH --account=[budget code]

# Load the module to make the Cray MPICH ABI available
module load cray-mpich-abi

export OMP_NUM_THREADS=1
export SRUN_CPUS_PER_TASK=$SLURM_CPUS_PER_TASK

# Set the LD_LIBRARY_PATH environment variable within the Singularity container
# to ensure that it used the correct MPI libraries.
export SINGULARITYENV_LD_LIBRARY_PATH="/opt/cray/pe/mpich/8.1.23/ofi/gnu/9.1/lib-abi-mpich:/opt/cray/pe/mpich/8.1.23/gtl/lib:/opt/cray/libfabric/1.12.1.2.2.0.0/lib64:/opt/cray/pe/gcc-libs:/opt/cray/pe/gcc-libs:/opt/cray/pe/lib64:/opt/cray/pe/lib64:/opt/cray/xpmem/default/lib64:/usr/lib64/libibverbs:/usr/lib64:/usr/lib64"

# This makes sure HPE Cray Slingshot interconnect libraries are available
# from inside the container.
export SINGULARITY_BIND="/opt/cray,/var/spool,/opt/cray/pe/mpich/8.1.23/ofi/gnu/9.1/lib-abi-mpich:/opt/cray/pe/mpich/8.1.23/gtl/lib,/etc/host.conf,/etc/libibverbs.d/mlx5.driver,/etc/libnl/classid,/etc/resolv.conf,/opt/cray/libfabric/1.12.1.2.2.0.0/lib64/libfabric.so.1,/opt/cray/pe/gcc-libs/libatomic.so.1,/opt/cray/pe/gcc-libs/libgcc_s.so.1,/opt/cray/pe/gcc-libs/libgfortran.so.5,/opt/cray/pe/gcc-libs/libquadmath.so.0,/opt/cray/pe/lib64/libpals.so.0,/opt/cray/pe/lib64/libpmi2.so.0,/opt/cray/pe/lib64/libpmi.so.0,/opt/cray/xpmem/default/lib64/libxpmem.so.0,/run/munge/munge.socket.2,/usr/lib64/libibverbs/libmlx5-rdmav34.so,/usr/lib64/libibverbs.so.1,/usr/lib64/libkeyutils.so.1,/usr/lib64/liblnetconfig.so.4,/usr/lib64/liblustreapi.so,/usr/lib64/libmunge.so.2,/usr/lib64/libnl-3.so.200,/usr/lib64/libnl-genl-3.so.200,/usr/lib64/libnl-route-3.so.200,/usr/lib64/librdmacm.so.1,/usr/lib64/libyaml-0.so.2"

# Launch the parallel job.
srun --hint=nomultithread --distribution=block:block \
    singularity exec ${EPCC_SINGULARITY_DIR}/osu_benchmarks.sif \
        osu_latency

Finally, submit the job to the batch system with

remote$ sbatch submit.slurm

Solution

As you can see in the job script shown above, we have called srun on the host system and are passing to MPI the singularity executable for which the parameters are the image file and the name of the benchmark executable we want to run.

The following shows an example of the output you should expect to see. You should have latency values reported for message sizes up to 4MB.

# OSU MPI Latency Test v5.6.2
# Size          Latency (us)
0                       0.38
1                       0.34
...

This has demonstrated that we can successfully run a parallel MPI executable from within a Singularity container.

Investigate performance of native benchmark compared to containerised version

To get an idea of any difference in performance between the code within your Singularity image and the same code built natively on the target HPC platform, try running the osu_allreduce benchmarks natively on ARCHER2 on all cores on at least 16 nodes (if you want to use more than 32 nodes, you will need to use the standard QoS rather than the short QoS). Then try running the same benchmark that you ran via the Singularity container. Do you see any performance differences?

What do you see?

Do you see the same when you run on small node counts - particularly a single node?

Note: a native version of the OSU micro-benchmark suite is available on ARCHER2 via module load osu-benchmarks.

Discussion

Here are some selected results measured on ARCHER2:

1 node:

  • 4 B
    • Native: 6.13 us
    • Container: 5.30 us (16% faster)
  • 128 KiB
    • Native: 173.00 us
    • Container: 230.38 us (25% slower)
  • 1 MiB
    • Native: 1291.18 us
    • Container: 2101.02 us (39% slower)

16 nodes:

  • 4 B
    • Native: 17.66 us
    • Container: 18.15 us (3% slower)
  • 128 KiB
    • Native: 237.29 us
    • Container: 303.92 us (22% slower)
  • 1 MiB
    • Native: 1501.25 us
    • Container: 2359.11 us (36% slower)

32 nodes:

  • 4 B
    • Native: 30.72 us
    • Container: 24.41 us (20% faster)
  • 128 KiB
    • Native: 265.36 us
    • Container: 363.58 us (26% slower)
  • 1 MiB
    • Native: 1520.58 us
    • Container: 2429.24 us (36% slower)

For the medium and large messages, using a container produces substantially worse MPI performance for this benchmark on ARCHER2. When the messages are very small, containers match the native performance and can actually be faster.

Is this true for other MPI benchmarks that use all the cores on a node or is it specific to Allreduce?

Summary

Singularity can be combined with MPI to create portable containers that run software in parallel across multiple compute nodes. However, there are some limitations, specifically:

Key Points

  • Singularity images containing MPI applications can be built on one platform and then run on another (e.g. an HPC cluster) if the two platforms have compatible MPI implementations.

  • When running an MPI application within a Singularity container, use the MPI executable on the host system to launch a Singularity container for each process.

  • Think about parallel application performance requirements and how where you build/run your image may affect that.

Break

Overview

Teaching: min
Exercises: min
Questions
Objectives

Lunch break

Key Points

Containers in Research Workflows: Reproducibility and Granularity

Overview

Teaching: 20 min
Exercises: 5 min
Questions

  • How can I use container images to make my research more reproducible?

  • How do I incorporate containers into my research workflow?

Objectives

  • Understand how container images can help make research more reproducible.

  • Understand what practical steps I can take to improve the reproducibility of my research using containers.

Although this workshop is titled “Reproducible computational environments using containers”, so far we have mostly covered the mechanics of using Singularity with only passing reference to the reproducibility aspects. In this section, we discuss these aspects in more detail.

Work in progress…

Note that reproducibility aspects of software and containers are an active area of research, discussion and development so are subject to many changes. We will present some ideas and approaches here but best practices will likely evolve in the near future.

Reproducibility

By reproducibility here we mean the ability of someone else (or your future self) being able to reproduce what you did computationally at a particular time (be this in research, analysis or something else) as closely as possible even if they do not have access to exactly the same hardware resources that you had when you did the original work.

Some examples of why containers are an attractive technology to help with reproducibility include:

Sharing images

We have made use of a few different online repositories during this course, such as Sylabs Cloud Library and Docker Hub which provide platforms for sharing container images publicly. Once you have uploaded a container image, you can point people to its public location and they can download and build upon it.

This is fine for working collaboratively with container images on a day-to-day basis but these repositories are not a good option for long time archive of container images in support of research and publications as:

Archiving and persistently identifying container images using Zenodo

When you publish your work or make it publicly available in some way it is good practice to make container images that you used for computational work available in an immutable, persistent way and to have an identifier that allows people to cite and give you credit for the work you have done. Zenodo is one service that provides this functionality.

Zenodo supports the upload of zip archives and we can capture our Singularity container images as zip archives. For example, to convert the container image we created earlier, alpine-sum.sif in this lesson to a zip archive (on the command line):

zip alpine-sum.zip alpine-sum.sif

Note: These zip container images can become quite large and Zenodo supports uploads up to 50GB. If your container image is too large, you may need to look at other options to archive them or work to reduce the size of the container images.

Once you have your archive, you can deposit it on Zenodo and this will:

In addition to the archive file itself, the deposit process will ask you to provide some basic metadata to classify the container image and the associated work.

Note that Zenodo is not the only option for archiving and generating persistent DOIs for container images. There are other services out there – for example, some organizations may provide their own, equivalent, service.

Reproducibility good practice

Container Granularity

As mentioned above, one of the decisions you may need to make when containerising your research workflows is what level of granularity you wish to employ. The two extremes of this decision could be characterized as:

Of course, many real applications will sit somewhere between these two extremes.

Positives and negatives

What are the advantages and disadvantages of the two approaches to container granularity for research workflows described above? Think about this and write a few bullet points for advantages and disadvantages for each approach in the course Etherpad.

Solution

This is not an exhaustive list but some of the advantages and disadvantages could be:

Single large container image

  • Advantages:
    • Simpler to document
    • Full set of requirements packaged in one place
    • Potentially easier to maintain (though could be opposite if working with large, distributed group)
  • Disadvantages:
    • Could get very large in size, making it more difficult to distribute
    • May end up with same dependency issues within the container image from different software requirements
    • Potentially more complex to test
    • Less re-useable for different, but related, work

Multiple smaller container images

  • Advantages:
    • Individual components can be re-used for different, but related, work
    • Individual parts are smaller in size making them easier to distribute
    • Avoid dependency issues between different pieces of software
    • Easier to test
  • Disadvantage:
    • More difficult to document
    • Potentially more difficult to maintain (though could be easier if working with large, distributed group)
    • May end up with dependency issues between component container images if they get out of sync

Next steps with containers

Now that we’re at the end of the lesson material, take a moment to reflect on what you’ve learned, how it applies to you, and what to do next.

  1. In your own notes, write down or diagram your understanding of Singularity containers and container images: concepts, commands, and how they work.
  2. In your own notes, write down how you think you might use containers in your daily work. If there’s something you want to try doing with containers right away, what is a next step after this workshop to make that happen?

Key Points

  • Container images allow us to encapsulate the computation (and data) we have used in our research.

  • Using online containerimage repositories allows us to easily share computational work we have done.

  • Using container images along with a DOI service such as Zenodo allows us to capture our work and enables reproducibility.

Additional topics and next steps

Overview

Teaching: 0 min
Exercises: 0 min
Questions

  • How do I understand more on how containers work?

  • What different container technologies are there and what are differences/implications?

  • How can I orchestrate different containers?

Objectives

  • Understand container technologies better.

  • Provide useful links to continue your journey with containers.

Additional topics

Key Points

  • TBC

(Optional) Using Singularity to run BLAST+

Overview

Teaching: 30 min
Exercises: 30 min
Questions

  • How can I use Singularity to run bioinformatics workflows with BLAST+?

Objectives

  • Show example of using Singularity with a common bioinformatics tool.

We have now learned enough to be able to use Sigularity to deploy software without us needed to install the software itself on the host system.

In this section we will demonstrate the use of a Singularity container image that provides the BLAST+ software.

Source material

This example is based on the example from the official NCBI BLAST+ Docker container documentation Note: the efetch parts of the step-by-step guide do not currently work using Singularity version of the image so we provide a dataset with the data already downloaded.

(This is because the NCBI BLAST+ Docker container image has the efetch tool installed in the /root directory and this special location gets overwritten during the conversion to a Singularity container image.)

Download the required data

Download the blast_example.tar.gz.

Unpack the archive which contains the downloaded data required for the BLAST+ example:

remote$ wget https://epcced.github.io/2023-09-21_Singularity_Nottingham/files/blast_example.tar.gz
remote$ tar -xvf blast_example.tar.gz

x blast/
x blast/blastdb/
x blast/queries/
x blast/fasta/
x blast/results/
x blast/blastdb_custom/
x blast/fasta/nurse-shark-proteins.fsa
x blast/queries/P01349.fsa

Finally, move into the newly created directory:

remote$ cd blast
remote$ ls  

blastdb        blastdb_custom fasta          queries        results

Create the Singularity container image

NCBI provide official Docker containers with the BLAST+ software hosted on Docker Hub. We can create a Singularity container image from the Docker container image with:

remote$ singularity pull ncbi-blast.sif docker://ncbi/blast

INFO:    Converting OCI blobs to SIF format
INFO:    Starting build...
Getting image source signatures
Copying blob f3b81f6693c5 done  
Copying blob 9e3ea8720c6d done  
Copying blob f1910abb61ed done  
Copying blob 5ac33d4de47b done  
Copying blob 8402427c8382 done  
Copying blob 06add1a477bc done  
Copying blob d9781f222125 done  
Copying blob 4aae31cc8a8b done  
Copying blob 6a61413c1ffa done  
Copying blob c657bf8fc6ca done  
Copying blob 1776e565f5f8 done  
Copying blob d90474a0d8c8 done  
Copying blob 0bc89cb1b9d7 done  
Copying blob b8a272fccf13 done  
Copying blob 891eb09f891f done  
Copying blob 4c64befa8a35 done  
Copying blob 7ab0b7afbc21 done  
Copying blob b007c620c60b done  
Copying blob f877ffc04713 done  
Copying blob 6ee97c348001 done  
Copying blob 03f0ee97190b done  
Copying config 28914b3519 done  
Writing manifest to image destination
Storing signatures
2023/06/16 08:26:53  info unpack layer: sha256:9e3ea8720c6de96cc9ad544dddc695a3ab73f5581c5d954e0504cc4f80fb5e5c
2023/06/16 08:26:53  info unpack layer: sha256:06add1a477bcffec8bac0529923aa8ae25d51f0660f0c8ef658e66aa89ac82c2
2023/06/16 08:26:53  info unpack layer: sha256:f3b81f6693c592ab94c8ebff2109dc60464d7220578331c39972407ef7b9e5ec
2023/06/16 08:26:53  info unpack layer: sha256:5ac33d4de47beb37ae35e9cad976d27afa514ab8cbc66e0e60c828a98e7531f4
2023/06/16 08:27:03  info unpack layer: sha256:8402427c8382ab723ac504155561fb6d3e5ea1e7b4f3deac8449cec9e44ae65a
2023/06/16 08:27:03  info unpack layer: sha256:f1910abb61edef8947e9b5556ec756fd989fa13f329ac503417728bf3b0bae5e
2023/06/16 08:27:03  info unpack layer: sha256:d9781f222125b5ad192d0df0b59570f75b797b2ab1dc0d82064c1b6cead04840
2023/06/16 08:27:03  info unpack layer: sha256:4aae31cc8a8b726dce085e4e2dc4671a9be28162b8d4e1b1c00b8754f14e6fe6
2023/06/16 08:27:03  info unpack layer: sha256:6a61413c1ffa309d92931265a5b0ecc9448568f13ccf3920e16aaacc8fdfc671
2023/06/16 08:27:03  info unpack layer: sha256:c657bf8fc6cae341e3835cb101dc4c6839ba4aad69578ff8538b3c1eba7abb21
2023/06/16 08:27:04  info unpack layer: sha256:1776e565f5f85562b8601edfd29c35f3fba76eb53177c8e89105f709387e3627
2023/06/16 08:27:04  info unpack layer: sha256:d90474a0d8c8e6165d909cc0ebbf97dbe70fd759a93eff11a5a3f91fa09a470e
2023/06/16 08:27:04  warn rootless{root/edirect/aux/lib/perl5/Mozilla/CA/cacert.pem} ignoring (usually) harmless EPERM on setxattr "user.rootlesscontainers"
2023/06/16 08:27:04  warn rootless{root/edirect/aux/lib/perl5/Mozilla/CA.pm} ignoring (usually) harmless EPERM on setxattr "user.rootlesscontainers"
2023/06/16 08:27:04  warn rootless{root/edirect/aux/lib/perl5/Mozilla/mk-ca-bundle.pl} ignoring (usually) harmless EPERM on setxattr "user.rootlesscontainers"
2023/06/16 08:27:04  info unpack layer: sha256:0bc89cb1b9d7ca198a7a1b95258006560feffaff858509be8eb7388b315b9cf5
2023/06/16 08:27:04  info unpack layer: sha256:b8a272fccf13b721fa68826f17f0c2bb395de377e0d22c98d38748eb5957a4c6
2023/06/16 08:27:04  info unpack layer: sha256:891eb09f891ff2c26f24a5466112e134f6fb30bd3d0e78c14c0d676b0e68d60a
2023/06/16 08:27:04  info unpack layer: sha256:4c64befa8a35c9f8518324524dfc27966753462a4c07b2234811865387058bf4
2023/06/16 08:27:04  info unpack layer: sha256:7ab0b7afbc21b75697a7b8ed907ee9b81e5b17a04895dc6ff7d25ea2ba1eeba4
2023/06/16 08:27:04  info unpack layer: sha256:b007c620c60b91ce6a9e76584ecc4bc062c822822c204d8c2b1c8668193d44d1
2023/06/16 08:27:04  info unpack layer: sha256:f877ffc04713a03dffd995f540ee13b65f426b350cdc8c5f1e20c290de129571
2023/06/16 08:27:04  info unpack layer: sha256:6ee97c348001fca7c98e56f02b787ce5e91d8cc7af7c7f96810a9ecf4a833504
2023/06/16 08:27:04  info unpack layer: sha256:03f0ee97190baebded2f82136bad72239254175c567b19def105b755247b0193
INFO:    Creating SIF file...

Now we have a container with the software in, we can use it.

Build and verify the BLAST database

Our example dataset has already downloaded the query and database sequences. We first use these downloaded data to create a custom BLAST database by using a container to run the command makeblastdb with the correct options.

remote$ singularity exec ncbi-blast.sif \
    makeblastdb -in fasta/nurse-shark-proteins.fsa -dbtype prot \
    -parse_seqids -out nurse-shark-proteins -title "Nurse shark proteins" \
    -taxid 7801 -blastdb_version 5


Building a new DB, current time: 06/16/2023 14:35:07
New DB name:   /home/auser/test/blast/blast/nurse-shark-proteins
New DB title:  Nurse shark proteins
Sequence type: Protein
Keep MBits: T
Maximum file size: 3000000000B
Adding sequences from FASTA; added 7 sequences in 0.0199499 seconds.

To verify the newly created BLAST database above, you can run the blastdbcmd -entry all -db nurse-shark-proteins -outfmt "%a %l %T" command to display the accessions, sequence length, and common name of the sequences in the database.

remote$ singularity exec ncbi-blast.sif \
    blastdbcmd -entry all -db nurse-shark-proteins -outfmt "%a %l %T"

Q90523.1 106 7801
P80049.1 132 7801
P83981.1 53 7801
P83977.1 95 7801
P83984.1 190 7801
P83985.1 195 7801
P27950.1 151 7801

Now we have our database we can run queries against it.

Run a query against the BLAST database

Lets execute a query on our database using the blastp command:

remote$ singularity exec ncbi-blast.sif \
    blastp -query queries/P01349.fsa -db nurse-shark-proteins \
    -out results/blastp.out

At this point, you should see the results of the query in the output file results/blastp.out. To view the content of this output file, use the command less results/blastp.out.

remote$ less results/blastp.out

...output trimmed...

Query= sp|P01349.2|RELX_CARTA RecName: Full=Relaxin; Contains: RecName:
Full=Relaxin B chain; Contains: RecName: Full=Relaxin A chain

Length=44
                                                                      Score     E
Sequences producing significant alignments:                          (Bits)  Value

P80049.1 RecName: Full=Fatty acid-binding protein, liver; AltName...  14.2    0.96


>P80049.1 RecName: Full=Fatty acid-binding protein, liver; AltName: Full=Liver-type
fatty acid-binding protein; Short=L-FABP
Length=132

...output trimmed...

With your query, BLAST identified the protein sequence P80049.1 as a match with a score of 14.2 and an E-value of 0.96.

Accessing online BLAST databases

As well as building your own local database to query, you can also access databases that are available online. For example, to see which databases are available online in the Google Compute Platform (GCP):

remote$ singularity exec ncbi-blast.sif update_blastdb.pl --showall pretty --source gcp

Connected to GCP
BLASTDB                                                      DESCRIPTION                                                                                                              SIZE (GB)      LAST_UPDATED
nr                                                           All non-redundant GenBank CDS translations+PDB+SwissProt+PIR+PRF excluding environmental samples from WGS projects        369.4824      2023-06-10
swissprot                                                    Non-redundant UniProtKB/SwissProt sequences                                                                                 0.3576      2023-06-10
refseq_protein                                               NCBI Protein Reference Sequences                                                                                          146.5088      2023-06-12
landmark                                                     Landmark database for SmartBLAST                                                                                            0.3817      2023-04-25
pdbaa                                                        PDB protein database                                                                                                        0.1967      2023-06-10
nt                                                           Nucleotide collection (nt)                                                                                                319.5044      2023-06-11
pdbnt                                                        PDB nucleotide database                                                                                                     0.0145      2023-06-09
patnt                                                        Nucleotide sequences derived from the Patent division of GenBank                                                           15.7342      2023-06-09
refseq_rna                                                   NCBI Transcript Reference Sequences                                                                                        47.8721      2023-06-12

...output trimmed...

Similarly, for databases hosted at NCBI:

remote$ singularity exec ncbi-blast.sif update_blastdb.pl --showall pretty --source ncbi

Connected to NCBI
BLASTDB                                                      DESCRIPTION                                                                                                              SIZE (GB)      LAST_UPDATED
env_nr                                                       Proteins from WGS metagenomic projects (env_nr).                                                                            3.9459      2023-06-11
SSU_eukaryote_rRNA                                           Small subunit ribosomal nucleic acid for Eukaryotes                                                                         0.0063      2022-12-05
LSU_prokaryote_rRNA                                          Large subunit ribosomal nucleic acid for Prokaryotes                                                                        0.0041      2022-12-05
16S_ribosomal_RNA                                            16S ribosomal RNA (Bacteria and Archaea type strains)                                                                       0.0178      2023-06-16
env_nt                                                       environmental samples                                                                                                      48.8599      2023-06-08
LSU_eukaryote_rRNA                                           Large subunit ribosomal nucleic acid for Eukaryotes                                                                         0.0053      2022-12-05
ITS_RefSeq_Fungi                                             Internal transcribed spacer region (ITS) from Fungi type and reference material                                             0.0067      2022-10-28
Betacoronavirus                                              Betacoronavirus                                                                                                            55.3705      2023-06-16

...output trimmed...

Notes

You have now completed a simple example of using a complex piece of bioinformatics software through Singularity containers. You may have noticed that some things just worked without you needing to set them up even though you were running using containers:

  1. We did not need to explicitly bind any files/directories in to the container. This worked because Singularity automatically binds the current directory into the running container, so any data in the current directory (or its subdirectories) will generally be available in running Singularity containers. (If you have used Docker containers, you will notice that this is different from the default behaviour there.)
  2. Access to the internet is automatically available within the running container in the same way as it is on the host system without us needed to specify any additional options.
  3. Files and data we create within the container have the right ownership and permissions for us to access outside the container.

In addition, we were able to use the tools in the container image provided by NCBI without having to do any work to install the software irrespective of the computing platform that we are using. (In fact, the example this is based on runs the pipeline using Docker on a cloud computing platform rather than on your local system.)

Key Points

  • We can use containers to run software without having to install it

  • The commands we use are very similar to those we would use natively

  • Singularity handles a lot of complexity around data and internet access for us