Software on Sapelo2 - Revision history

Shtsai at 17:45, 27 March 2026

2026-03-27T17:45:20Z

← Older revision		Revision as of 13:45, 27 March 2026
Line 1:		Line 1:
	=Introduction=		==Introduction==

	On Sapelo2 users have the option to [https://wiki.gacrc.uga.edu/wiki/Installing_Applications_on_Sapelo2 install their own software] or use software installed centrally on Sapelo2 by the GACRC staff. Centrally-installed software on Sapelo2 will typically be in one of two formats: a software module or a Singularity container. Outlined here is an explanation of each of both categories of software and how to use them.		On Sapelo2 users have the option to [https://wiki.gacrc.uga.edu/wiki/Installing_Applications_on_Sapelo2 install their own software] or use software installed centrally on Sapelo2 by the GACRC staff. Centrally-installed software on Sapelo2 will typically be in one of two formats: a software module or a Singularity container. Outlined here is an explanation of each of both categories of software and how to use them.

	=Getting Help Using Software=		==Getting Help Using Software==

	Very often the best place to learn how to use various types of scientific software will be the software's official documentation. Some software may have its own website, whereas other software's documentation may be in a README file of a Github repository. This is where you will find information such as a how to call software, necessary input files, default parameters values, various options available when calling software, etc... This type of documentation is very important to read to make sure that you're using software as you intend to and to its fullest extent.		Very often the best place to learn how to use various types of scientific software will be the software's official documentation. Some software may have its own website, whereas other software's documentation may be in a README file of a Github repository. This is where you will find information such as a how to call software, necessary input files, default parameters values, various options available when calling software, etc... This type of documentation is very important to read to make sure that you're using software as you intend to and to its fullest extent.
Line 40:		Line 40:
	bc06026@ra4-2 ~$ bowtie2</nowiki>		bc06026@ra4-2 ~$ bowtie2</nowiki>

	=Software Modules=		==Software Modules==

	The majority of software centrally installed on Sapelo2 is installed in the form of a '''software module'''. A software module is a grouping of some software and dependencies. By leveraging software modules, you gain access to only the software you need when you need it. This is achieved by modifying your PATH environmental variable as well as creating other environment variables.		The majority of software centrally installed on Sapelo2 is installed in the form of a '''software module'''. A software module is a grouping of some software and dependencies. By leveraging software modules, you gain access to only the software you need when you need it. This is achieved by modifying your PATH environmental variable as well as creating other environment variables.
Line 206:		Line 206:
	In the above example we see that Python/3.10.4-GCCcore-11.3.0 was loaded, and then there was an attempt to import and print the version of SciPy from the command line. This returned an error, because SciPy has its own software module on Sapelo2, which includes Python. We can see that after loading the SciPy-bundle/2022.05-foss-2022a module, the same command printed out 1.8.1 for the version of SciPy.		In the above example we see that Python/3.10.4-GCCcore-11.3.0 was loaded, and then there was an attempt to import and print the version of SciPy from the command line. This returned an error, because SciPy has its own software module on Sapelo2, which includes Python. We can see that after loading the SciPy-bundle/2022.05-foss-2022a module, the same command printed out 1.8.1 for the version of SciPy.

	=Singularity Containers=		==Singularity Containers==

	Some software on Sapelo2 is installed in the form of a Singularity container, in /apps/singularity-images. Singularity is an open-source container technology, similar to Docker, but designed for HPC cluster environments. For more information on Singularity, please see their documentation [https://sylabs.io/docs/ here]. Please note that Docker cannot be run on Sapelo2 due to security considerations of Docker containers giving a user effectively root privileges. Docker also requires a daemon, whereas Singularity does not. The good news is that you can convert Docker images into Singularity images.		Some software on Sapelo2 is installed in the form of a Singularity container, in /apps/singularity-images. Singularity is an open-source container technology, similar to Docker, but designed for HPC cluster environments. For more information on Singularity, please see their documentation [https://sylabs.io/docs/ here]. Please note that Docker cannot be run on Sapelo2 due to security considerations of Docker containers giving a user effectively root privileges. Docker also requires a daemon, whereas Singularity does not. The good news is that you can convert Docker images into Singularity images.

Shtsai at 21:08, 9 January 2026

2026-01-09T21:08:21Z

← Older revision		Revision as of 17:08, 9 January 2026
Line 10:		Line 10:

	<nowiki>		<nowiki>
	bc06026@ra4-2 ~$ ml Bowtie2/2.4.5-GCC-11.3.0		bc06026@ra4-2 ~$ ml Bowtie2/2.5.4-GCC-13.3.0
	bc06026@ra4-2 ~$		bc06026@ra4-2 ~$
	bc06026@ra4-2 ~$ bowtie2 --help		bc06026@ra4-2 ~$ bowtie2 --help
	Bowtie 2 version 2.4.5 by Ben Langmead (langmea@cs.jhu.edu, www.cs.jhu.edu/~langmea)		Bowtie 2 version 2.5.4 by Ben Langmead (langmea@cs.jhu.edu, www.cs.jhu.edu/~langmea)
	Usage:		Usage:
	bowtie2 [options]* -x <bt2-idx> {-1 <m1> -2 <m2> \| -U <r> \| --interleaved <i> \| -b <bam>} [-S <sam>]		bowtie2 [options]* -x <bt2-idx> {-1 <m1> -2 <m2> \| -U <r> \| --interleaved <i> \| -b <bam>} [-S <sam>]
Line 31:		Line 31:
	-h/--help print this usage message		-h/--help print this usage message
	</nowiki>		</nowiki>
	In the above help output for the Bowtie2 software, we see example usage, and some of the optional parameters that you can specify when calling bowtie2. If you're not sure what the name of the compiled binary to call is, it very often will be the name of the software you're using in lowercase (but not always). You can verify this by checking online documentation, or by loading the software and then starting to type what may be the appropriate executable, and then hitting the tab key twice on your keyboard, which would auto-complete the rest of the executable name if it is valid. Hitting tab three times in succession would show you if there are any similarly-named executables of the software with the same name. For example, this output is shown after loading the Bowtie2/2.4.5-GCC-11.3.0 module and typing "bow" and then hitting tab three times:		In the above help output for the Bowtie2 software, we see example usage, and some of the optional parameters that you can specify when calling bowtie2. If you're not sure what the name of the compiled binary to call is, it very often will be the name of the software you're using in lowercase (but not always). You can verify this by checking online documentation, or by loading the software and then starting to type what may be the appropriate executable, and then hitting the tab key twice on your keyboard, which would auto-complete the rest of the executable name if it is valid. Hitting tab three times in succession would show you if there are any similarly-named executables of the software with the same name. For example, this output is shown after loading the Bowtie2/2.5.4-GCC-13.3.0 module and typing "bow" and then hitting tab three times:

	<nowiki>		<nowiki>

Shtsai: /* Software Modules */

2025-08-01T15:51:15Z

Software Modules

← Older revision		Revision as of 11:51, 1 August 2025
Line 48:		Line 48:
	Here are some examples of Python modules on Sapelo2:		Here are some examples of Python modules on Sapelo2:

	<nowiki>Python/2.7.18-GCCcore-11.3.0		<nowiki>Python/2.7.18-GCCcore-12.3.0
	Python/3.9.6-GCCcore-11.2.0		Python/3.10.4-GCCcore-11.3.0
	Python/3.10.4-GCCcore-11.3.0</nowiki>		Python/3.11.3-GCCcore-12.3.0
			Python/3.12.3-GCCcore-13.3.0</nowiki>

	Here is an example of how the path and version of the command "python" changes after a module is loaded:		Here is an example of how the path and version of the command "python" changes after a module is loaded:
Line 59:		Line 60:
	/usr/bin/python3		/usr/bin/python3
	bc06026@ra4-2 ~$python3 -V		bc06026@ra4-2 ~$python3 -V
	Python 3.6.8		Python 3.9.21
	bc06026@ra4-2 ~$ ml Python/3.10.4-GCCcore-11.3.0		bc06026@ra4-2 ~$ ml Python/3.10.4-GCCcore-11.3.0
	bc06026@ra4-2 ~$ which python		bc06026@ra4-2 ~$ which python
Line 146:		Line 147:
	Python is a programming language that lets you work more quickly and integrate your systems more effectively.		Python is a programming language that lets you work more quickly and integrate your systems more effectively.

	Versions:		Versions:
	Python/2.7.18-GCCcore-11.2.0		Python/2.7.18-GCCcore-11.3.0-bare
	Python/2.7.18-GCCcore-11.3.0		Python/2.7.18-GCCcore-12.3.0
	Python/3.7.4-GCCcore-8.3.0		Python/2.7.18-GCCcore-13.3.0
	Python/3.8.6-GCCcore-10.2.0		Python/3.9.6-GCCcore-11.2.0-bare
	Python/3.9.6-GCCcore-11.2.0		Python/3.9.6-GCCcore-11.2.0
			Python/3.10.4-GCCcore-11.3.0-bare
	Python/3.10.4-GCCcore-11.3.0		Python/3.10.4-GCCcore-11.3.0
			Python/3.11.3-GCCcore-12.3.0
			Python/3.11.5-GCCcore-13.2.0
			Python/3.12.3-GCCcore-13.3.0
	Other possible modules matches:		Other possible modules matches:
	Biopython Boost.Python IPython bx-python ~~netcdf4~~-python ~~openslide~~-python		Biopython Boost.Python Brotli-python GitPython IPython Python-bundle-PyPI bx-python flatbuffers-python graphviz-python intervaltree-python ...

	-------------------------------------------------------------------------------------------------------------------		-------------------------------------------------------------------------------------------------------------------
Line 167:		Line 171:
	For example:		For example:

	$ module spider Python/3.10.4-GCCcore-11.3.0		$ module spider Python/3.12.3-GCCcore-13.3.0
	-------------------------------------------------------------------------------------------------------------------		-------------------------------------------------------------------------------------------------------------------

Shtsai: /* Getting Help Using Software */

2025-08-01T15:47:30Z

Getting Help Using Software

← Older revision		Revision as of 11:47, 1 August 2025
Line 7:		Line 7:
	Very often the best place to learn how to use various types of scientific software will be the software's official documentation. Some software may have its own website, whereas other software's documentation may be in a README file of a Github repository. This is where you will find information such as a how to call software, necessary input files, default parameters values, various options available when calling software, etc... This type of documentation is very important to read to make sure that you're using software as you intend to and to its fullest extent.		Very often the best place to learn how to use various types of scientific software will be the software's official documentation. Some software may have its own website, whereas other software's documentation may be in a README file of a Github repository. This is where you will find information such as a how to call software, necessary input files, default parameters values, various options available when calling software, etc... This type of documentation is very important to read to make sure that you're using software as you intend to and to its fullest extent.

	When the online documentation for software is lacking or non-existent, you may also be able to find some helpful information calling the software via the command line. To do this, first start an interactive session with <code>~~qlogin~~</code>. In your interactive session, load the software you want to use (be it a software module, Conda environment, or Singularity container) and then execute the relevant compiled binary or script with no options, or with -h or --help. For example:		When the online documentation for software is lacking or non-existent, you may also be able to find some helpful information calling the software via the command line. To do this, first start an interactive session with <code>interact</code>. In your interactive session, load the software you want to use (be it a software module, Conda environment, or Singularity container) and then execute the relevant compiled binary or script with no options, or with -h or --help. For example:

	<nowiki>		<nowiki>

Shtsai: /* Building Singularity Containers */

2025-05-29T18:04:04Z

Building Singularity Containers

← Older revision		Revision as of 14:04, 29 May 2025
Line 256:		Line 256:


	'''NOTE''': By default the cache directory for the singularity pull command ~~will be~~ in the user's home directory, in $HOME/.apptainer/cache. This cache directory can become very large and take up a lot of space in this hidden (.apptainer) directory. We recommend that users set the singularity cache directory to be in their /scratch directory, by defining e.g.		'''NOTE''': By default the cache directory for the singularity pull command is in the user's home directory, in $HOME/.apptainer/cache. This cache directory can become very large and take up a lot of space in this hidden (.apptainer) directory. We recommend that users set the singularity cache directory to be in their /scratch directory, by defining e.g.
	<pre class="gcommand">		<pre class="gcommand">
	export APPTAINER_CACHE=/scratch/$USER/apptainer		export APPTAINER_CACHE=/scratch/$USER/apptainer

Shtsai: /* Building Singularity Containers */

2025-05-29T18:03:41Z

Building Singularity Containers

← Older revision		Revision as of 14:03, 29 May 2025
Line 255:		Line 255:
	This will download a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.		This will download a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.

	~~Please note that by~~ default the cache directory for the singularity pull command will be in the user's home directory, in $HOME/.apptainer/cache. This cache directory can become very large and take up a lot of space in this hidden (.apptainer) directory. We recommend that users set the singularity cache directory to be in their /scratch directory, by defining e.g.
			'''NOTE''': By default the cache directory for the singularity pull command will be in the user's home directory, in $HOME/.apptainer/cache. This cache directory can become very large and take up a lot of space in this hidden (.apptainer) directory. We recommend that users set the singularity cache directory to be in their /scratch directory, by defining e.g.
	<pre class="gcommand">		<pre class="gcommand">
	export APPTAINER_CACHE=/scratch/$USER/apptainer		export APPTAINER_CACHE=/scratch/$USER/apptainer

Shtsai at 18:03, 29 May 2025

2025-05-29T18:03:00Z

← Older revision		Revision as of 14:03, 29 May 2025
Line 254:		Line 254:

	This will download a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.		This will download a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.

			Please note that by default the cache directory for the singularity pull command will be in the user's home directory, in $HOME/.apptainer/cache. This cache directory can become very large and take up a lot of space in this hidden (.apptainer) directory. We recommend that users set the singularity cache directory to be in their /scratch directory, by defining e.g.
			<pre class="gcommand">
			export APPTAINER_CACHE=/scratch/$USER/apptainer
			</pre>
			This way the singularity cache will not contribute towards the user's home dir quota and those cache files will eventually be cleared by the scratch purge policy, even if the user does not delete them once the singularity container is created.

	<!--		<!--
	Regardless of whether you build a Singularity image this way with a definition file or a Docker image, the Singularity image that you create will also exist at the URL provided at the end of the image build output, albeit with a tokenized name. Clicking on one of your images there will take you to a webpage that provides a <code>singularity pull</code> command for that Singularity image. If you would prefer to build a Singularity image in your web browser, you also have the option of writing or uploading a definition file to https://cloud.sylabs.io/builder. This also requires having and signing into a cloud.sylabs.io account, but provides an in-browser editor to write your own Singularity definition files, or to upload one you've already written.		Regardless of whether you build a Singularity image this way with a definition file or a Docker image, the Singularity image that you create will also exist at the URL provided at the end of the image build output, albeit with a tokenized name. Clicking on one of your images there will take you to a webpage that provides a <code>singularity pull</code> command for that Singularity image. If you would prefer to build a Singularity image in your web browser, you also have the option of writing or uploading a definition file to https://cloud.sylabs.io/builder. This also requires having and signing into a cloud.sylabs.io account, but provides an in-browser editor to write your own Singularity definition files, or to upload one you've already written.

Shtsai at 15:13, 19 August 2024

2024-08-19T15:13:16Z

← Older revision		Revision as of 11:13, 19 August 2024
Line 286:		Line 286:
	* If the documentation of the software you want to use is lacking and you're unable to determine what executable binary or script is meant to be run from the Singularity image you're using, try using the <code>ls</code> command to search directories inside the container. It is very common for a container's software to be in /usr/bin or /opt.		* If the documentation of the software you want to use is lacking and you're unable to determine what executable binary or script is meant to be run from the Singularity image you're using, try using the <code>ls</code> command to search directories inside the container. It is very common for a container's software to be in /usr/bin or /opt.
	* Singularity containers have been configured to access to the user's home directory ($HOME), scratch directory (/scratch), and the local scratch directory on the node (/lscratch). The /tmp directory is defined inside the container.		* Singularity containers have been configured to access to the user's home directory ($HOME), scratch directory (/scratch), and the local scratch directory on the node (/lscratch). The /tmp directory is defined inside the container.
			* To run a GPU-enabled singularity container on the GPU, please submit the job to the gpu_p partition, request a GPU device and add the --nv option to the singularity command.

Shtsai at 19:00, 31 July 2024

2024-07-31T19:00:37Z

← Older revision		Revision as of 15:00, 31 July 2024
Line 206:		Line 206:
	Some software on Sapelo2 is installed in the form of a Singularity container, in /apps/singularity-images. Singularity is an open-source container technology, similar to Docker, but designed for HPC cluster environments. For more information on Singularity, please see their documentation [https://sylabs.io/docs/ here]. Please note that Docker cannot be run on Sapelo2 due to security considerations of Docker containers giving a user effectively root privileges. Docker also requires a daemon, whereas Singularity does not. The good news is that you can convert Docker images into Singularity images.		Some software on Sapelo2 is installed in the form of a Singularity container, in /apps/singularity-images. Singularity is an open-source container technology, similar to Docker, but designed for HPC cluster environments. For more information on Singularity, please see their documentation [https://sylabs.io/docs/ here]. Please note that Docker cannot be run on Sapelo2 due to security considerations of Docker containers giving a user effectively root privileges. Docker also requires a daemon, whereas Singularity does not. The good news is that you can convert Docker images into Singularity images.

	Singularity images are run to create Singularity containers. An image will be either a .sif or .simg (old extension name) file on Sapelo2. From Singularity's [https://sylabs.io/docs/ documentation], an image is "a single executable file based container image, cryptographically signed, auditable, secure, and easy to move using existing data mobility paradigms." Like a Docker image, this is essentially an executable program to start a container. From Singularity's [https://singularity.lbl.gov/#:~:text=Singularity%20enables%20users%20to%20have,a%20Singularity%20container%20and%20run. documentation], "Singularity containers can be used to package entire scientific workflows, software and libraries, and even data." Like a Docker container, a Singularity container is a user space apart from the underlying operating system from where you started the container. This typically contain any necessary dependencies for the application, and in some cases its own operating system in the container. An example where this can be very convenient is if an application was designed to run in an Ubuntu environment, but you need to run it on Sapelo2, which is ~~CentOS~~. By using a Singularity container, you are able to create the necessary environment for your software in your job, whether interactive or through a submission script.		Singularity images are run to create Singularity containers. An image will be either a .sif or .simg (old extension name) file on Sapelo2. From Singularity's [https://sylabs.io/docs/ documentation], an image is "a single executable file based container image, cryptographically signed, auditable, secure, and easy to move using existing data mobility paradigms." Like a Docker image, this is essentially an executable program to start a container. From Singularity's [https://singularity.lbl.gov/#:~:text=Singularity%20enables%20users%20to%20have,a%20Singularity%20container%20and%20run. documentation], "Singularity containers can be used to package entire scientific workflows, software and libraries, and even data." Like a Docker container, a Singularity container is a user space apart from the underlying operating system from where you started the container. This typically contain any necessary dependencies for the application, and in some cases its own operating system in the container. An example where this can be very convenient is if an application was designed to run in an Ubuntu environment, but you need to run it on Sapelo2, which is Rocky Linux. By using a Singularity container, you are able to create the necessary environment for your software in your job, whether interactive or through a submission script.

	===Singularity Commands===		===Singularity Commands===
Line 216:		Line 216:
	* <code>singularity shell</code> - Launches a Singularity container and starts an interactive shell inside the container.		* <code>singularity shell</code> - Launches a Singularity container and starts an interactive shell inside the container.
	<!--<code>singularity build --remote</code> - Builds a Singularity container remotely, and then pulls the container on to the machine from which this command was run. This is a great feature, because without the --remote option, building Singularity containers requires administrative privileges.-->		<!--<code>singularity build --remote</code> - Builds a Singularity container remotely, and then pulls the container on to the machine from which this command was run. This is a great feature, because without the --remote option, building Singularity containers requires administrative privileges.-->

			'''Please note:''' You can run <code>singularity</code> on the compute nodes, either in an interactive session or in a batch job, without loading any modules. Note that singularity cannot be run on the Sapelo2 login nodes.

	===Building Singularity Containers===		===Building Singularity Containers===

Moses: /* Building Singularity Containers */

2023-10-10T16:06:11Z

Building Singularity Containers

← Older revision		Revision as of 12:06, 10 October 2023
Line 248:		Line 248:

	* Building a Singularity image by pulling a Docker image from Docker Hub:		* Building a Singularity image by pulling a Docker image from Docker Hub:

	<code>apptainer pull docker://user/image:tag</code>		<code>apptainer pull docker://user/image:tag</code>

	This will ~~pull~~ a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.		This will download a Docker image and convert it to a Singularity image, and then put it to the local machine from which the build command was ran.
	<!--		<!--
	Regardless of whether you build a Singularity image this way with a definition file or a Docker image, the Singularity image that you create will also exist at the URL provided at the end of the image build output, albeit with a tokenized name. Clicking on one of your images there will take you to a webpage that provides a <code>singularity pull</code> command for that Singularity image. If you would prefer to build a Singularity image in your web browser, you also have the option of writing or uploading a definition file to https://cloud.sylabs.io/builder. This also requires having and signing into a cloud.sylabs.io account, but provides an in-browser editor to write your own Singularity definition files, or to upload one you've already written.		Regardless of whether you build a Singularity image this way with a definition file or a Docker image, the Singularity image that you create will also exist at the URL provided at the end of the image build output, albeit with a tokenized name. Clicking on one of your images there will take you to a webpage that provides a <code>singularity pull</code> command for that Singularity image. If you would prefer to build a Singularity image in your web browser, you also have the option of writing or uploading a definition file to https://cloud.sylabs.io/builder. This also requires having and signing into a cloud.sylabs.io account, but provides an in-browser editor to write your own Singularity definition files, or to upload one you've already written.