In order for the scripts in these examples to work, you will need to replace <ACCOUNT> with your group's account name.
Hello World
This script will print "Hello World", sleep for 10 seconds, and then print the time and date. The output will be written to a file in your current directory.
#!/bin/sh # # Simple "Hello World" submit script for Slurm. # # Replace <ACCOUNT> with your account name before submitting. # #SBATCH --account=<ACCOUNT> # The account name for the job. #SBATCH --job-name=HelloWorld # The job name. #SBATCH -c 1 # The number of cpu cores to use. #SBATCH --time=1:00 # The time the job will take to run. #SBATCH --mem-per-cpu=1gb # The memory the job will use per cpu core. echo "Hello World" sleep 10 date # End of script
C/C++/Fortran
To submit a precompiled binary to run on Terremoto, the script will look just as it does in the Hello World example. The difference is that you will call your executable file instead of the shell commands "echo", "sleep", and "date".
C/C++/Fortran MPI
Intel Parallel Studio
Terremoto supports Intel Parallel Studio which provides a version of MPI derived from MPICH2. We encourage users to avail themselves of Intel MPI because it is faster and more modern than other versions. Also, all nodes on the cluster have Infiniband transport and that is the fabric that MPI jobs avail themselves of - which is another reason for a substantial boost of efficiency on the cluster.
To use Intel MPI, you must load the Intel module first:
module load intel-parallel-studio/2017 mpiexec -bootstrap slurm ./myprogram
In order to take advantage of Terremoto architecture, your program should be (re)compiled on the cluster even if you used Intel for compiling it on another cluster. It is important to compile with the compiler provided by the module mentioned above. Note that you may have to set additional environment variables in order to successfully compile your program.
These are the locations of the C and Fortran compilers for Intel Studio:
$ module load intel-parallel-studio/2017 (...) $ which mpiicc /moto/opt/parallel_studio_xe_2017/compilers_and_libraries_2017.0.098/linux/mpi/intel64/bin/mpiicc $ which ifort /moto/opt/parallel_studio_xe_2017/compilers_and_libraries_2017.0.098/linux/bin/intel64/ifort
For programs written in C, use mpiicc in order to compile them:
$ mpiicc -o <MPI_OUTFILE> <MPI_INFILE.c>
The submit script below, named pi_mpi.sh, assumes that you have compiled a simple MPI program used to compute pi, (see mpi_test.c), and created a binary called pi_mpi:
#!/bin/sh #SBATCH -A <ACCOUNT> #SBATCH --time=30 #SBATCH -N 2 #SBATCH --exclusive module load intel-parallel-studio/2017 mpiexec -bootstrap slurm ./pi_mpi # End of script
The --exclusive flag will ensure that full nodes are being used in the runs (that's the reason why no memory specification is given). Each available core will give rise to another MPI thread. Without the flag, you can specify the number of tasks, or tasks per node, in order to limit the number of threads that will be created. For example, you can replace the directive containing the flag by:
#SBATCH -N 2 #SBATCH --ntasks-per-node=4
- and your MPI code will run on 8 threads, with 4 on each of the 2 nodes requested.
Job Submission
$ sbatch pi_mpi.sh
OpenMPI
Terremoto supports also OpenMPI from the GNU family.
To use OpenMPI, you must load the following module instead:
module load openmpi/gcc/64 mpiexec myprogram
Your program must be compiled on the cluster. You can use the the module command as explained above to set your path so that the corresponding mpicc will be found. Note that you may have to set additional environment variables in order to successfully compile your program.
$ module load openmpi/gcc/64 $ which mpicc /moto/opt/openmpi-2.0.1/bin/mpicc
Compile your program using mpicc. For programs written in C:
$ mpicc -o <MPI_OUTFILE> <MPI_INFILE.c>
GPU (CUDA C/C++)
The cluster includes 8 Nvidia V100 GPU servers each with 2 GPU modules per server.
To use a GPU server you must specify the --gres=gpu option in your submit request, followed by a colon and the number of GPU modules you require (with a maximum of 2 per server).
Request a v100 gpu, specify this in your submit script.
#SBATCH --gres=gpu
Not all applications have GPU support, but some, such as MATLAB, have built-in GPU support and can be configured to use GPUs.
To build your CUDA code and run it on the GPU modules you must first set your paths so that the Nvidia compiler can be found. Please note you must be logged into a GPU node to access these commands. To login interactively to a GPU node, run the following command, replacing <ACCOUNT> with your account.
$ srun --pty -t 0-01:00 --gres=gpu:1 -A <ACCOUNT> /bin/bash
Load the cuda environment module which will add cuda to your PATH and set related environment variables. Note cuda 8.0 does not support gcc 6, so gcc 5 or earlier must be accessible in your environment when running nvcc.
$ module load gcc/4.8.5
Load the cuda module.
$ module load cuda92/toolkit
You then have to compile your program using nvcc:
$ nvcc -o <EXECUTABLE_NAME> <FILE_NAME.cu>
You can compile hello_world.cu sample code which can be built with the following command:
$ nvcc -o hello_world hello_world.cu
For non-trivial code samples, refer to Nvidia's CUDA Toolkit Documentation.
A Slurm script template, gpu.sh, that can be used to submit this job is shown below:
#!/bin/sh # #SBATCH --account=<ACCOUNT> # The account name for the job. #SBATCH --job-name=HelloWorld # The job name. #SBATCH --gres=gpu:1 # Request 1 gpu (Up to 4 on K80s, or up to 2 on P100s are valid). #SBATCH -c 1 # The number of cpu cores to use. #SBATCH --time=1:00 # The time the job will take to run. #SBATCH --mem-per-cpu=1gb # The memory the job will use per cpu core. module load cuda92/toolkit ./hello_world # End of script
Job submission
$ sbatch gpu.sh
This program will print out "Hello World!" when run on a gpu server or print "Hello Hello" when no gpu module is found.
Singularity
Singularity is a software tool that brings Docker-like containers and reproducibility to scientific computing and HPC. Singularity has Docker container support and enables users to easily run different flavors of Linux with different software stacks. These containers provide a single universal on-ramp from the laptop, to HPC, to cloud.
Users can run Singularity containers just as they run any other program on our HPC clusters. Example usage of Singularity is listed below. For additional details on how to use Singularity, please contact us or refer to the Singularity User Guide.
Downloading Pre-Built Containers
Singularity makes it easy to quickly deploy and use software stacks or new versions of software. Since Singularity has Docker support, users can simply pull existing Docker images from Docker Hub or download docker images directly from software repositories that increasingly support the Docker format. Singularity Container Library also provides a number of additional containers.
You can use the pull command to download pre-built images from an external resource into your current working directory. The docker:// uri reference can be used to pull Docker images. Pulled Docker images will be automatically converted to the Singularity container format.
This example pulls the default Ubuntu docker image from docker hub.
$ singularity pull docker://ubuntu
Running Singularity Containers
Here's an example of pulling the latest stable release of the Tensorflow Docker image and running it with Singularity. (Note: these pre-built versions may not be optimized for use with our CPUs.)
First, load the Singularity software into your environment with:
$ module load singularity
Then pull the docker image. This also converts the downloaded docker image to Singularity format and save it in your current working directory:
$ singularity pull docker://tensorflow/tensorflow
Done. Container is at: ./tensorflow.simgOnce you have download a container, you can run it interactively in a shell or in batch mode.
Singularity - Interactive Shell
The shell command allows you to spawn a new shell within your container and interact with it as though it were a small virtual machine:
$ singularity shell tensorflow.simg
Singularity: Invoking an interactive shell within container...From within the Singularity shell, you will see the Singularity prompt and can run the downloaded software. In this example, python is launched and tensorflow is loaded.
Singularity tensorflow.simg:~> python >>> import tensorflow as tf >>> print(tf.__version__) 1.13.1 >>> exit()
When done, you may exit the Singularity interactive shell with the "exit" command.
Singularity tensorflow.simg:~> exit
Singularity: Executing Commands
The exec command allows you to execute a custom command within a container by specifying the image file. This is the way to invoke commands in your job submission script.
$ module load singularity
$ singularity exec tensorflow.simg [command]
For example, to run python example above using the exec command:
$ singularity exec tensorflow.simg python -c 'import tensorflow as tf; print(tf.__version__)'
Singularity: Running a Batch Job
Below is an example of job submission script named submit.sh that runs Singularity. Note that you may need to specify the full path to the Singularity image you wish to run.
#!/bin/bash # Singularity example submit script for Slurm. # # Replace <ACCOUNT> with your account name before submitting. # #SBATCH -A <ACCOUNT> # Set Account name #SBATCH --job-name=tensorflow # The job name #SBATCH -c 1 # Number of cores #SBATCH -t 0-0:30 # Runtime in D-HH:MM #SBATCH --mem-per-cpu=4gb # Memory per cpu core module load singularity singularity exec tensorflow.simg python -c 'import tensorflow as tf; print(tf.__version__)'
Then submit the job to the scheduler. This example prints out the tensorflow version.
$ sbatch submit.sh
For additional details on how to use Singularity, please contact us or refer to the Singularity User Guide.
Example of R run
For this example, the R code below is used to generate a graph ''Rplot.pdf'' of a discrete Delta-hedging of a call. It hedges along a path and repeats over many paths. There are two R files required:
A Slurm script, hedge.sh, that can be used to submit this job is presented below:
#!/bin/sh #hedge.sh #Slurm script to run R program that generates graph of discrete Delta-hedging call #Slurm directives # #SBATCH -A astro # The account name for the job. #SBATCH -J DeltaHedge # The job name. #SBATCH -c 6 # The number of cpu cores to use. #SBATCH -t 1:00 # The time the job will take to run. #SBATCH --mem-per-cpu 1gb # The memory the job will use per cpu core. module load R #Command to execute R code R CMD BATCH --no-save --vanilla hedge.R routput # End of script
Batch queue submission
$ sbatch hedge.sh
This program will leave several files in the output directory: slurm-<jobid>.out, Rplots.pdf, and routput (the first one will be empty).
Installing R Packages on Terremoto
HPC users can Install R packages locally in their home directory or group's scratch space (see below).
Local Installation
After logging in to Terremoto, start R:
$ module load R $ R
You can see the default library paths (where R looks for packages) by calling .libPaths():
> .libPaths() [1] "/moto/opt/R-3.5.1/lib64/R/library"
These paths are all read-only, and so you cannot install packages to them. To fix this, we will tell R to look in additional places for packages.
Exit R and create a directory rpackages in /moto/<GROUP>/users/<UNI>/.
$ mkdir /moto/<GROUP>/users/<UNI>/rpackages
Go back into R and add this path to .libPaths()
$ R
> .libPaths("/moto/<GROUP>/users/<UNI>/rpackages/")
Call .libPaths() to make sure the path has been added
> .libPaths() [1] "/moto/<GROUP>/users/<UNI>/rpackages/" [2] "/usr/lib64/R/site-library" [3] "/usr/lib64/R/library"
To install a package, such as the "sm" package, tell R to put the package in your newly created local library:
> install.packages("sm", lib="/moto/<GROUP>/users/<UNI>/rpackages")
Select appropriate mirror and follow install instructions.
Test to see if package can be called:
> library(sm) Package `sm', version 2.2-3; Copyright (C) 1997, 2000, 2005, 2007 A.W.Bowman & A.Azzalinitype help(sm) for summary information
In order to access this library from your programs, make sure you add the following line to the top of every program:
.libPaths("/moto/<GROUP>/users/<UNI>/rpackages/")
Since R will know where to look for libraries, a call to library(sm) will be successful (however, this line is not necessary per se for the install.packages(...) call, as the directory is already specified in it).
Matlab
Matlab (single thread)
The file linked below is a Matlab M-file containing a single function, simPoissGLM, that takes one argument (lambda).
A Slurm script, simpoiss.sh, that can be used to submit this job is presented below (implicitly, --cpu-per-task=1).
#!/bin/sh # # Simple Matlab submit script for Slurm. # # #SBATCH -A astro # The account name for the job. #SBATCH -J SimpleMLJob # The job name. #SBATCH -t 1:00 # The time the job will take to run. #SBATCH --mem-per-cpu=1gb # The memory the job will use per cpu core. module load matlab echo "Launching an Matlab run" date #define parameter lambda LAMBDA=10 #Command to execute Matlab code matlab -nosplash -nodisplay -nodesktop -r "simPoissGLM($LAMBDA)" # > matoutfile # End of script
Batch queue submission
$ sbatch simpoiss.sh
This program will leave several files in the output directory: slurm-<jobid>.out, out.mat, and matoutfile.
Matlab (multi-threading)
Matlab has built-in implicit multi-threading (even without applying its Parallel Computing Toolbox, PCT), which causes it to use several cores on the node it is running on. It consumes the number of cores assigned by Slurm.The user can activate explicit (PCT) multi-threading by specifying the number of cores desired also in the Matlab program.
The Torque submit script (simpoiss.sh) should contain the following line:
#SBATCH -c 6
The -c flag determines the number of cores (up to 24 are allowed).
For explicit multi-threading, the users must include the following corresponding statement within their Matlab program:
parpool('local', 6)
The second argument passed to parpool must equal the number specified with the ppn directive. Users who are acquainted with the use of commands like parfor need to specify explicit multi-threading with the help of parpool command above.
Note: maxNumCompThreads() is being deprecated by Mathworks. It is being replaced by parpool:
The command to execute Matlab code remains unchanged from the single thread example above.
Important note: On Yeti, where Matlab was single thread by default, it appeared that the more recent versions of Matlab took liberties to grab all the cores within a node even when fewer (or even only one) cores were specified as above. On Terremoto, we believe this has been addressed by implementing a system mechanism which enforces the proper usage of the number of specified cores.
Matlab with Parallel Server
Matlab 2020b and 2022b on Terremoto now have access to Parallel Server, and the toolbox is installed. The first time you run Matlab, it can take a few minutes to fully open, especially over WiFi. In order to use Parallel Server, a Cluster Profile needs to be created to use the Slurm job scheduler. You will need to request the number of nodes desired as well and may need to increase the amount of memory desired. With an interactive job requesting two nodes start with:
|
Step One
Using the Configure for Slurm MathWorks tutorial as a guide:
- On the Home tab, in the Environment area, select Parallel > Create and Manage Clusters. Click ok on the dialog box Product Required: MATLAB Parallel Server.
- Create a new profile in the Cluster Profile Manager by selecting Add Cluster Profile > Slurm.
- With the new profile selected in the list, click Rename and edit the profile name something informative for future use, e.g., InstallTest. Press Enter.
- In the Properties tab, provide settings for the following fields:
- Set the Description field to something informative, e.g., For testing installation.
- Set the JobStorageLocation to the location where you want job and task data to be stored, e.g.,
/moto/home/<your-directory>.- Note: JobStorageLocation should not be shared by parallel computing products running different versions; each version on your cluster should have its own JobStorageLocation.
- Set the NumWorkers field to the number of workers you want to run the validation tests on. This should be not be more than what is specified by
--nodes=in the interactive job request, i.e.,srun. - Set the ClusterMatlabRoot to the installation location of the MATLAB version, i.e.,
/moto/opt/matlab/R2020bor/moto/opt/matlab/R2022b. - Within ADDITIONAL SLURM PROPERTIES add
- Click Done to save your cluster profile.
Step 2: Validate the Cluster Profile
Step Two
In this step you verify your cluster profile, and thereby your installation. You can specify the number of workers to use when validating your profile. If you do not specify the number of workers in the Validation tab, then the validation will attempt to use as many workers as the value specified by the NumWorkers property on the Properties tab. You can specify a smaller number of workers to validate your configuration without occupying the whole cluster.
- If it is not already open, start the Cluster Profile Manager from the MATLAB desktop. On the Home tab, in the Environment area, select Parallel > Create and Manage Clusters.
- Select your cluster profile in the listing.
- Click Validation tab.
- Use the checkboxes to choose all tests, or a subset of the validation stages, and specify the number of workers to use when validating your profile.
- Click Validate. Note when the Parallel pool test (
parpool) starts running, the screen flips back to Matlab, and in the very bottom left status bar, you will see Starting Parallel Pool on the profile name you created in Step 1. - The Validation Results tab shows the output as shown in the MathWorks tutorial.
- If your validation passed, you now have a valid profile that you can use in other parallel applications. You can make any modifications to your profile appropriate for your applications, such as NumWorkersRange, AttachedFiles, AdditionalPaths, etc.
Python and JULIA
To use python you need to use:
$ module load anaconda
Here's a simple python program called "example.py" – it has just one line:
print ("Hello, World!")
To submit it on the Terremoto Cluster, use the submit script "example.sh"
(*** use "astro" if you are a member of "astro" group, otherwise use your group name):
#!/bin/sh # # Simple "Hello World" submit script for Slurm. # #SBATCH --account=astro # The account name for the job. #SBATCH --job-name=HelloWorld # The job name. #SBATCH -c 1 # The number of cpu cores to use. #SBATCH --time=1:00 # The time the job will take to run. #SBATCH --mem-per-cpu=1gb # The memory the job will use per cpu core. module load anaconda #Command to execute Python program python example.py #End of script
If you use "ls" command you should see 2 programs:
example.sh example.py
To submit it - please use:
$ sbatch example.sh
To check the output use:
$ cat slurm-463023.out Hello, World!
Similarly, here is the "julia_example.jl" with just one line
$ cat julia_example.jl
println("hello world")
and
$ cat julia_example.sh
#!/bin/sh # # Simple "Hello World" submit script for Slurm. # #SBATCH --account=hblab # The account name for the job. #SBATCH --job-name=HelloWorld # The job name. #SBATCH -c 1 # The number of cpu cores to use. #SBATCH --time=1:00 # The time the job will take to run. #SBATCH --mem-per-cpu=1gb # The memory the job will use per cpu core. module load julia #Command to execute Python program julia julia_example.jl #End of script
After you finish creating those two files, if you use "ls"command you should see:
julia_example.jl julia_example.sh
To submit it use:
$ sbatch julia_example.sh Submitted batch job 463030
To check the output
$ cat slurm-463030.out hello world
Julia Interactive Session Usage:
Step 1 >> start an interactive session (*** use "astro" if you are a member of "astro" group, otherwise use your group name):
$ srun --pty -t 0-04:00 -A astro /bin/bash $ module load julia $ julia julia_example.jl hello world $ julia _ _ _ _(_)_ | A fresh approach to technical computing () | () (_) | Documentation: http://docs.julialang.org _ _ _| |_ __ _ | Type "?help" for help. | | | | | | |/ _` | | | | || | | | (| | | _/ |_'|||_'| | Official http://julialang.org/ release |__/ | x86_64-pc-linux-gnu julia> To quit Julia use "CTRL +D"
Julia packages can be installed with this command (for example "DataFrames" package):
julia> using Pkg
julia> Pkg.add("DataFrames")
Please check this website:
https://pkg.julialang.org/docs/
to see the full list of the official packages available.
Tensorflow
Tensorflow computations can use CPUs or GPUs. The default is to use CPUs which are more prevalent, but typically slower than GPUs.
Tensorflow for CPUs
Tensorflow (optimized for Terremoto CPUs) and Keras are available by loading the anaconda/3-2018.12 module:
$ module load anaconda/3-2018.12
$ python
Python 3.7.1 (default, Dec 14 2018, 19:28:38)
>>> import tensorflow as tf
>>> print(tf.__version__)
1.13.1
>>> import keras
Using TensorFlow backend.
>>> print(keras.__version__)
2.2.4
>>> exit()
Tensorflow with GPU Support
The following describes how you to run tensorflow on a Terremoto GPU node. GPUs will typically run Tensorflow computations much faster than CPUs.
First, run an interactive job requesting one GPU on a GPU node
$ srun --pty -t 0-02:00:00 --gres=gpu:1 -A <group_name> /bin/bash
Then load the singularity environment module and run the tensorflow container, which was built from the Tensorflow docker image. You can start an interactive singularity shell and specify the --nv flag which instructs singularity to use the Nvidia GPU driver.
$ module load singularity
$ singularity shell --nv /moto/opt/singularity/tensorflow-1.13-gpu-py3-moto.simg
Singularity tensorflow-1.13-gpu-py3-moto.simg:~> python
Python 3.5.2 (default, Nov 12 2018, 13:43:14)
[GCC 5.4.0 20160609] on linux
>>> import tensorflow as tf
>>> hello = tf.constant('Hello, TensorFlow!')
>>> sess = tf.Session()
..
>>> exit()
You may type "exit" to exit when you're done with the Singularity shell.
Singularity tensorflow-1.13-gpu-py3-moto.simg:~> exit
Below is an example of job submission script named submit.sh that runs Tensorflow with GPU support using Singularity.
#!/bin/bash # Tensorflow with GPU support example submit script for Slurm. # # Replace <ACCOUNT> with your account name before submitting. # #SBATCH -A <ACCOUNT> # Set Account name #SBATCH --job-name=tensorflow # The job name #SBATCH -c 1 # Number of cores #SBATCH -t 0-0:30 # Runtime in D-HH:MM #SBATCH --gres=gpu:1 # Request a gpu module module load singularity singularity exec --nv /moto/opt/singularity/tensorflow-1.13-gpu-py3-moto.simg python -c 'import tensorflow as tf; print(tf.__version__)'
Then submit the job to the scheduler. This example prints out the tensorflow version.
$ sbatch submit.sh
For additional details on how to use Singularity, please contact us, see our Singularity documentation, or refer to the Singularity User Guide.
Another option:
Please note that you should not work on our head node.
Since we have a limited number of GPU nodes, it is not unusual to wait in queue to get a GPU node.
You should request an interactive job on GPU node (specify your groups name as accountName):
$ srun --pty -t 0-02:00:00 --gres=gpu:1 -A <accountNAME> /bin/bash
$ module load tensorflow/anaconda3-2019.10/gpu-2.0
Start python
$ python
Python 3.7.4 (default, Aug 13 2019, 20:35:49)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
To test it:
>>> import tensorflow as tf
>>> print(tf.__version__)
2.0.0
Jupyter Notebooks
This is one way to set up and run a jupyter notebook on Terremoto. As your notebook will listen on a port that will be accessible to anyone logged in on a submit node you should first create a password.
Creating a Password
The following steps can be run on the submit node or in an interactive job.
1. Load the anaconda python module.
$ module load anaconda/3-2019.10
2. If you haven’t already done so, initialize your jupyter environment.
$ jupyter notebook --generate-config
3. Start a python or ipython session.
$ ipython
4. Run the password hash generator. You will be prompted for a password, prompted again to verify, and then a hash of that password will be displayed.
In [1]: from notebook.auth import passwd; passwd() Enter password: Verify password: Out[1]: 'sha1:60bdb1:306fe0101ca73be2429edbab0935c545'
5. Cut and paste the hash into ~/.jupyter/jupyter_notebook_config.py
(Important: the following line in the file is commented out by default so please uncomment it first)
c.NotebookApp.password = 'sha1:60bdb1:306fe0101ca73be2429edbab0935c545'
Setting the password will prevent other users from having access to your notebook and potentially causing confusion.
Running a Jupyter Notebook
1. Log in to the submit node. Start an interactive job.
$ srun --pty -t 0-01:00 -A <ACCOUNT> /bin/bash
Please note that the example above specifies time limit of one hour only. That can be set to a much higher value, and in fact the default (i.e. if not specified at all) is as long as 5 days.
2. Get rid of XDG_RUNTIME_DIR environment variable
$ unset XDG_RUNTIME_DIR
3. Load the anaconda environment module.
$ module load anaconda/3-2019.10
4. Look up the IP of the node your interactive job is running on.
$ hostname -i 10.43.4.206
5. Start the jupyter notebook, specifying the node IP.
$ jupyter notebook --no-browser --ip=10.43.4.206
6. Look for the following line in the startup output to get the port number.
The Jupyter Notebook is running at: http://10.43.4.206:8888/
7. From your local system, open a second connection to Terremoto that forwards a local port to the remote node and port. Replace UNI below with your uni.
$ ssh -L 8080:10.43.4.206:8888 UNI@moto.rcs.columbia.edu
8. Open a browser session on your desktop and enter the URL 'localhost:8080' (i.e. the string within the single quotes) into its search field. You should now see the notebook.