VSC/Slurm/Spack Tips and Tricks

published 2022-05-18 (last changed on 2024-03-14) by Lukas Winkler

This is not official documentation for the Vienna Scientific Cluster. For this check the VSC Wiki. Instead, this is my personal cheat sheet of things that are not well documented elsewhere. Also while the content is focused on the VSC, most of the things mentioned here also apply to similar setups that use Slurm at other universities.

# Basics

Always request an interactive session when running anything using a non-trivial amount of CPU power!

# Quick interactive session

$ salloc --ntasks=2 --mem=2G --time=01:00:00

Don’t forget to then connect to the node you get assigned:

$ ssh n1234-567

# Storage

official docs:

$HOME is limited to 100 GB and storing/compiling code. Anything else should be stored at $DATA.

# Quota

The file size and number of files is limited by group. The current status can be read using

$ mmlsquota --block-size auto -j data_fs00000 data

for $DATA and

$ mmlsquota --block-size auto -j home_fs00000 home

for $HOME where 00000 is the ID of the own project (accessible using groups)

# Job scripts

# Basic Template

Your job script is a regular bash script (.sh file). In addition, you can specify options to sbatch in the beginning of your file:

#SBATCH --job-name=somename
#SBATCH --mail-type=ALL

In these cases --long-option=value and --long-option value are equivalent.

# Single Core job

Only specify --ntask=1 and the amount of memory you need.

#SBATCH --ntasks=1 # (also -n 1)
#SBATCH --mem 2G

# More sbatch options

All options can be found in the slurm documentation. A few useful ones are:

# Useful Environment Variables

Especially the latter can be used e.g. for running MPI programs with the requested number of CPU cores:

$ mpiexec -np $SLURM_NPROCS ./program

# Submitting Jobs

A job script can be submitted using

$ sbatch # you can also add sbatch options here

Just like in regular shell scripts, you can pass arguments to like this

$ sbatch somevalue

and then access somevalue as $1 in your script. This way multiple similar jobs can be submitted without needing to edit the jobscript.

# Queue

The current status of jobs in the Queue can be seen using squeue.

$ squeue -u username

Especially useful is the estimated start time of a scheduled job:

$ squeue -u username --start

A lot more information about scheduling including the calculated priority of jobs can be found using sprio

$ sprio -u username

This will also show the reason why the job is still queued for which an explanation can be found in the slurm documentation or the VSC wiki.

Details about past Jobs (like maximum memory usage), can be found using sacct. You can manually specify the needed columns or display most of them using --long

$ sacct -j 2052157 --long 

# Watch job output

If you submit a job, the output of the jobscript will by default be written to slurm-000000.out in the same directory. If you want to watch the output as it is written to the file while the job is running, tail -f filename will watch the file. But as the output file is only created once the job starts, this might fail. Using tail -F instead avoids this issue by watching the filename instead, which detects once the file appears and then watches it:

$ tail -F slurm-952347.out 
tail: cannot open 'slurm-952347.out' for reading: No such file or directory
tail: 'slurm-952347.out' has appeared;  following new file
The output of your job

# Advanced Slurm features

# QoS, accounts and partitions

Depending on access to private nodes, you might have access to many different QoS (Quality of Service), accounts and partitions.

On VSC you can get an overview over your account with sqos (this is also shown on login):

# this only works on VSC
$ /opt/adm/vsc/cluster-admin/src/python_modules/vsc_node_tools/scripts/ account
Your jobs can run with the following account(s) and quality of service (QoS):

default_account:                       p12345 
        account:                       p12345 

    default_qos:                  p12345_0512 
            qos:                  p12345_0512 

If you want to a different account or QoS than your default (e.g. if you want to access private nodes or GPU nodes), you can specify them with --qos and --acccount in salloc, sbatch or your job script.

You can also get an overview over all available partitions with sinfo and specify one explicitly with --partition.

You can get a better overview over the current usage of the QoS using sqos. This gives you the number of CPUs that are currently in use out of the limit for each QoS:

# this only works on VSC
$ sqos
     qos name  type  total res   used res   free res        walltime   priority   total n*    used n*    free n* 
  p71867_0512   cpu       7680       5632       2048     10-00:00:00     100000         30         22          8 
* node values do not always align with resource values since nodes can be partially allocated

# Array Jobs

Sometimes you might want to submit a larger number of similar jobs. This can be easily achieved using array jobs and the --array argument. With this, your job will be submitted multiple times with a different task ID that can be used from the $SLURM_ARRAY_TASK_ID environment variable.

#SBATCH --array=0-26
./your_program $SLURM_ARRAY_TASK_ID
Keep in mind that each individual job should not be too small (more than just a few minutes) as otherwise the computational overhead of scheduling the job and starting it will not be worth it. In these cases using one job that runs the program in a loop will be more efficient.

# Ignore specific nodes

Sometimes it is possible that one node is broken in some way (maybe one of the GPUs is no longer detected), but your jobs still get assigned to it. Then you can use --exclude to exclude individual or ranges of nodes from your job.

$ sbatch --exclude n1234-[001-002],n1234-008

# SSH login via

official docs (but we are using the more modern ProxyJump instead of Agent forwarding as this way we don’t have to trust the intermediate server with our private key)

Access to VSC is only possible from IP addresses of the partner universities. If you are from the University of Vienna and don’t want to use the VPN, an SSH tunnel via is an alternative.

To connect to the login server, the easiest thing is to put the config for the host in your ~/.ssh/config (create it, if it doesn’t yet exist).

Host loginUnivie
    User testuser12 # replace with your username
    # the following are needed if you are using OpenSSH 8.8 or newer
    # and the login server isn't yet updated to a never version
    HostkeyAlgorithms +ssh-rsa
    PubkeyAcceptedAlgorithms +ssh-rsa

This way you should now be able to test connecting to the login server using

$ ssh loginUnivie

Then you can add another entry to ~/.ssh/config on your computer for VSC that uses ProxyJump to connect via the loginUnivie entry we just created.

Host vsc5
    User vscuser
    ProxyJump loginUnivie
    # Port 27 # (only use if you are using ssh keys)
$ ssh vsc5

# Spack Modules

(official docs, that this guide builds on. More useful tips can be found in the spack documentation)

Software that is needed can be loaded via modules. The easiest way to find the right module for the current processor architecture, is directly querying spack, which is used to provide all compiled libraries and applications. There should never be a need to run module directly and doing so might accidentally pick libraries that are not intended for the current processor architecture.

# Finding the right module

The easiest way is using spack find.

$ spack find cmake

In case this only returns one module that fits your requirements, you can directly replace spack find with spack load to load this module.

But most of the time, you will find multiple modules which differ in their properties (and spack load will fail if the query resolves to more than one package):

$ spack find cmake
-- linux-almalinux8-zen / gcc@8.5.0 -----------------------------

-- linux-almalinux8-zen2 / gcc@9.5.0 ----------------------------

-- linux-almalinux8-zen3 / aocc@4.0.0 ---------------------------

-- linux-almalinux8-zen3 / gcc@12.2.0 ---------------------------

-- linux-almalinux8-zen3 / intel@2021.7.1 -----------------------
==> 5 installed packages

The most important property is the version and it is denoted with an @ sign. Another property is the compiler the program or library was compiled with and it can be separated with a % (and an additional @ for the version of the compiler).

So if you want to load e.g. cmake version 3.x.x compiled with gcc version 12, you could directly search for it and subsequently load it.

$ spack find cmake@3%gcc@12
$ spack load cmake@3%gcc@12 

This way if another minor update of cmake is released, your command will load it. If you don’t like this, check the next section.

Sometimes there are also multiple variants of the same module. spack info modulename can give you an overview over all of them, but that doesn’t mean that all combinations of variants/compilers/versions are offered at VSC. If you are for example interested in the hdf5 library with MPI support, you can search for the following (-v gives you the exact properties of each module):

$ spack find -v hdf5 +mpi

# “Locking” modules

If you dislike the fact that spack load queries don’t resolve to specific packages, but just filters that describe the properties you want or prefer exactly specifying the version of a package for reproducibility, you can find the hash of package using spack find -l and can then use /hash to always refer to this exact package:

$ spack find -l gsl%gcc@12
-- linux-almalinux8-zen3 / gcc@12.2.0 ---------------------------
whc7rma gsl@2.7.1
==> 1 installed package
$ spack load /whc7rma

# Find currently loaded modules

# List all currently loaded packages
$ spack find --loaded
# Unload all currently loaded packages
$ spack unload --all

# Libraries not found at runtime

Sometimes a program that just compiled without any issues (as the correct spack modules are loaded) won’t run afterwards as the libraries are not found at run time.

./your_program: error while loading shared libraries: cannot open shared object file: No such file or directory

This is caused by a recent change in Spack: $LD_LIBRARY_PATH is no longer set by default, to avoid loading a module breaking unrelated software. You can avoid this by setting $LD_LIBRARY_PATH to the value of $LIBRARY_PATH after loading your modules (as the latter is managed by spack).


Keep in mind that doing so might bring back the issues that changing $LD_LIBRARY_PATH causes.

# Comparing modules

Sometimes two packages look exactly the same:

$ spack find -vl fftw
-- linux-almalinux8-zen3 / intel@2021.9.0 -----------------------
ugklsjz fftw@3.3.10+mpi+openmp~pfft_patches build_system=autotools precision=double,float
j7dhzrq fftw@3.3.10+mpi+openmp~pfft_patches build_system=autotools precision=double,float

Then you can use spack diff to find the exact difference in them (most likely the modules that were used to compile this module)

$ spack diff /mmgor5w /cy5tkce
--- fftw@3.3.10/mmgor5w3daiwtsdbyl4dfhjsueaciry2
+++ fftw@3.3.10/cy5tkcetpgx35rok2lqfi3d66rjptkva
@@ depends_on @@
-  fftw intel-oneapi-mpi build
+  fftw openmpi build

Therefore, we know that in this example the first package depends on intel-oneapi-mpi and the second one on openmpi.

# Debugging modules

Sometimes one needs to know what spack load somepackage does exactly (e.g. because a library is still not found even though you loaded the module). Adding --sh to spack load prints out all commands that would be executed during the module load allowing you to understand what is going on.

$ spack load --sh cmake%gcc@12
export ACLOCAL_PATH=[...];
export CMAKE_PREFIX_PATH=[...];
export CPATH=[...];
export LD_RUN_PATH=[...];
export LIBRARY_PATH=[...];
export MANPATH=[...];
export PATH=[...];
export PKG_CONFIG_PATH=[...];

# Commonly used modules

This is a list of modules I commonly use. While it might not be directly usable for other people and will go out of date quickly, it might still serve as a good starting point.

spack load openmpi@4.1.6%gcc@12.2/exh7lqk
spack load --only package fftw@3.3%gcc@12.2/42q2cmu
spack load libtool%gcc@12.2 # GNU Autotools
spack load --only package hdf5%gcc@12.2/z3jjmoe # +mpi
spack load numactl%gcc@12.2
spack load metis%gcc@12.2
spack load intel-tbb%gcc@12.2
spack load gsl%gcc@12.2
spack load cmake@3.24%gcc@12.2
spack load gcc@12.2
spack load --only package python@3.11.3%gcc@12

# Examples

The examples in this section are not intended to be official documentation on how to use these software packages or the most efficient way to utilize VSC, but rather examples on how VSC usage can look in practice and inspiration for own jobscripts.


We are using MUSIC to generate initial simulations for cosmological simulations and then run them using SWIFT.

# Managing modules

$ cat 
spack load openmpi@4.1.6%gcc@12.2/exh7lqk
spack load --only package fftw@3.3%gcc@12.2/42q2cmu
spack load libtool%gcc@12.2 # GNU Autotools
spack load --only package hdf5%gcc@12.2/z3jjmoe # +mpi
spack load numactl%gcc@12.2
spack load metis%gcc@12.2
spack load intel-tbb%gcc@12.2
spack load gsl%gcc@12.2
spack load cmake@3.24%gcc@12.2
spack load gcc@12.2
$ source

# Compiling MUSIC

$ git clone
$ cd music/
$ mkdir build
$ cd build/
$ cmake ..
$ make -j 20

# Compiling SWIFT

$ git clone
$ cd swiftsim/
$ ./
$ ./
$ ./configure --enable-fof --enable-compiler-warnings=yes --enable-stand-alone-fof --with-numa --with-tbbmalloc 
# check the summary to see if everything you need is enabled/detected
$ make -j 20
$ ./swift # or ./swift_mpi

# A job-script for MUSIC

#SBATCH --job-name=music-auriga6
#SBATCH --mail-type=ALL
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --exclusive
#SBATCH --time=2:00:00


cd $ROOT
source ./

cd $ROOT/data/$run/

$MUSIC music.conf

Submit as sbatch something to run using data/something/music.conf.

# A job-script for small SWIFT simulations

For (very) small numbers of particles MPI overhead might slow down the simulation more than speed it up. Therefore, we could just use one of the CPUs of a shared node.

#SBATCH --job-name=auriga6
#SBATCH --mail-type=ALL
#SBATCH --ntasks=1
#SBATCH --cpus-per-task 64
#SBATCH --ntasks-per-node=1
#SBATCH --time=24:00:00
#SBATCH --mem=128GB


cd $ROOT
source ./

cd $ROOT/data/$run/

$swift --cosmology --self-gravity --fof --limiter --threads=64 --pin run_auriga6.yml

# A job-script for large SWIFT simulations

Check “Running on Large Systems” for more information.

#SBATCH --job-name=auriga6
#SBATCH --mail-type=ALL
#SBATCH --nodes=4
#SBATCH --ntasks-per-node=2
#SBATCH --exclusive
#SBATCH --time=24:00:00


cd $ROOT
source ./

cd $ROOT/data/$run/

mpirun -np 8 $swift --cosmology --self-gravity --fof --limiter --threads=64 --pin run_auriga6.yml

# Former guides

The following sections have been removed from the main guide as they are most likely no longer valid.

# Avoiding broken programs due to loaded dependencies

Recent versions of spack don’t set $LD_LIBRARY_PATH any more, which means that “unnecessarily” loaded spack modules should no longer affect other programs at runtime. If you manually modify $LD_LIBRARY_PATH you might still run into these issues now.

Loading a spack module not just loads the specified module, but also all dependencies of this module. With some modules like openmpi that dependency tree can be quite large.

$ spack find -d openmpi%gcc
-- linux-almalinux8-zen3 / gcc@11.2.0 ---------------------------

And loading module like openssl or ncurses from spack means that programs that depend on those libraries, but the versions provided by the base operating system, will crash.

$ spack load openmpi%gcc
$ nano somefile.txt
Segmentation fault (core dumped)
$ htop
Segmentation fault (core dumped)

One can avoid this by unloading the affected modules afterwards.

$ spack unload ncurses
$ spack unload openssl

But in many cases one doesn’t need all dependency modules and is really just interested in e.g. openmpi itself. Therefore, one can ignore the dependencies with --only package.

# doesn't affect non-openmpi programs
$ spack load --only package openmpi%gcc 

You have any feedback or ideas to improve this? Contact me per E-Mail or edit the source directly. You can find my other projects at