Submitting a job to a slurm queue

2025-05-26T14:16:31Z

Dvory: /* Basic Script */

== Accessing the System ==

To submit jobs to SLURM at Tel Aviv University, you need to access the system through one of the following login nodes:

* slurmlogin.tau.ac.il

=== Requirements for Access ===

* '''Group Membership''': You must be part of the "power" group to access the resources.
* '''University Credentials''': Use your Tel Aviv University username and password to log in.

These login nodes are your starting point for submitting jobs, checking job status, and managing your SLURM tasks.

=== SSH Example ===

To access the system using SSH, use the following example:

<syntaxhighlight lang="bash">
# Replace 'your_username' with your actual Tel Aviv University username
ssh your_username@slurmlogin.tau.ac.il
</syntaxhighlight>

Your connection will be automatically routed to one of the login nodes:
powerslurm-login, powerslurm-login2, or powerslurm-login3.

If you have an SSH key set up for password-less login, you can specify it like this:

<syntaxhighlight lang="bash">
# Replace 'your_username' and '/path/to/your/private_key' accordingly
ssh -i /path/to/your/private_key your_username@slurmlogin.tau.ac.il
</syntaxhighlight>

== Environment Modules ==

Environment Modules in SLURM allow users to dynamically modify their shell environment, providing an easy way to load and unload different software applications, libraries, and their dependencies. This system helps avoid conflicts between software versions and ensures the correct environment for running specific applications.

Here are some common commands to work with environment modules:<syntaxhighlight lang="bash">
#List Available Modules: To see all the modules available on the system, use:
module avail

#To search for a specific module by name (e.g., `gcc`), use:
module avail gcc/gcc-12.1.0

#Get Detailed Information About a Module: The `module spider` command provides detailed information about a module, including versions, dependencies, and descriptions:
module spider gcc/gcc-12.1.0

#View Module Settings: To see what environment variables and settings will be modified by a module, use:
module show gcc/gcc-12.1.0

#Load a Module: To set up the environment for a specific software, use the `module load` command. For example, to load GCC version 12.1.0:
module load gcc/gcc-12.1.0

#List Loaded Modules: To view all currently loaded modules in your session, use:
module list

#Unload a Module: To unload a specific module from your environment, use:
module unload gcc/gcc-12.1.0

#Unload All Modules:** If you need to clear your environment of all loaded modules, use:
module purge

</syntaxhighlight>By using these commands, you can easily manage the software environments needed for different tasks, ensuring compatibility and reducing potential conflicts between software versions.

== Basic Job Submission Commands ==

=== Finding Your Account and Partition ===

Before submitting a job, you need to know which partitions you have permission to use.

Run the command `<code>check_my_partitions</code>` to view a list of all the partitions you have permission to send jobs to.

== Submitting Jobs==
sbatch: Submits a job script for batch processing.

'''Example''':<syntaxhighlight lang="bash">

sbatch --ntasks=1 --time=10 -p power-general -A power-general-users pre_process.bash
# This command submits pre_process.bash to the power-general partition for 10 minutes.
# With 1 GPU:
sbatch --gres=gpu:1 -p gpu-general -A gpu-general-users gpu_job.sh

</syntaxhighlight>

=== Submitting Multiple Jobs ===

If you need to submit many similar jobs (hundreds or more), you should use a **Slurm job array**. Submitting each job individually using separate `sbatch` commands places a heavy load on the scheduler, slowing down job processing across the cluster. Job arrays allow you to bundle many related jobs together as a single submission. This is more efficient and easier to manage.

Each task in the array runs independently like a separate job, but the array is submitted as a single job ID for scheduling and tracking purposes.
You can customize the behavior of each task using the environment variable <code>SLURM_ARRAY_TASK_ID</code>.

==== Script Example: Job Array ====

This script submits a job array with 100 tasks, each processing a different input file. The array reduces scheduler load and simplifies job tracking.

<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=array_job # Job name
#SBATCH --account=power-general-users # Account name
#SBATCH --partition=power-general # Partition name
#SBATCH --time=02:00:00 # Max run time (hh:mm:ss)
#SBATCH --ntasks=1 # Number of tasks per array job
#SBATCH --nodes=1 # Number of nodes
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --mem-per-cpu=4G # Memory per CPU
#SBATCH --array=1-100 # Array range: 100 tasks
#SBATCH --output=array_job_%A_%a.out # Output file: Job ID and array task ID
#SBATCH --error=array_job_%A_%a.err # Error file: Job ID and array task ID

echo "Starting SLURM array task"
echo "Job ID: $SLURM_JOB_ID"
echo "Array Task ID: $SLURM_ARRAY_TASK_ID"
echo "Running on node(s): $SLURM_JOB_NODELIST"
echo "Allocated CPUs: $SLURM_JOB_CPUS_PER_NODE"

# Your application commands go here
# You can use $SLURM_ARRAY_TASK_ID to customize behavior per task
# ./my_program input_${SLURM_ARRAY_TASK_ID}.txt
echo "Task completed"
</syntaxhighlight>

In this example:
* The job array consists of 100 tasks.
* Each task runs the same script but with a different input file.
* You access the task ID using the environment variable <code>SLURM_ARRAY_TASK_ID</code>.
* The output and error logs are separated per task using <code>%A</code> (job ID) and <code>%a</code> (array task ID).

==== Script Example: Job Array with different parameters per task ====

This script submits a job array with 3 tasks. Each task runs the same program with a different input file: `data1.txt`, `data2.txt`, and `data3.txt`.

<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=array_job # Job name
#SBATCH --account=power-general-users # Account name
#SBATCH --partition=power-general # Partition name
#SBATCH --time=01:00:00 # Max run time (hh:mm:ss)
#SBATCH --ntasks=1 # Number of tasks per array job
#SBATCH --nodes=1 # Number of nodes
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --mem-per-cpu=2G # Memory per CPU
#SBATCH --array=1-3 # Run 3 tasks with IDs 1, 2, 3
#SBATCH --output=array_%A_%a.out # Output file: Job ID and task ID
#SBATCH --error=array_%A_%a.err # Error file: Job ID and task ID

echo "Starting SLURM array task"
echo "Job ID: $SLURM_JOB_ID"
echo "Array Task ID: $SLURM_ARRAY_TASK_ID"

# Each task runs the program with a different input file
./my_program data${SLURM_ARRAY_TASK_ID}.txt

echo "Task completed"
</syntaxhighlight>
===Writing Single SLURM Job Scripts===

Here is a simple job script example:

==== Basic Script====
<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=my_job # Job name
#SBATCH --account=power-general-users # Account name
#SBATCH --partition=power-general # Partition name
#SBATCH --time=02:00:00 # Max run time (hh:mm:ss)
#SBATCH --ntasks=1 # Number of tasks
#SBATCH --nodes=1 # Number of nodes
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --mem-per-cpu=4G # Memory per CPU
#SBATCH --output=my_job_%j.out # Output file
#SBATCH --error=my_job_%j.err # Error file
#SBATCH --mail-user=<your email> # Your mail address to receive an email
#SBATCH --mail-type=END,FAIL # The mail will be sent upon ending the script successfully or not

echo "Starting my SLURM job"
echo "Job ID: $SLURM_JOB_ID"
echo "Running on nodes: $SLURM_JOB_NODELIST"
echo "Allocated CPUs: $SLURM_JOB_CPUS_PER_NODE"

# Your application commands go here
# ./my_program

echo "Job completed"
</syntaxhighlight>

To ask for x cores interactively:
<pre>
srun --ntasks=1 --cpus-per-task=x --partition=power-general --nodes=1 --pty bash
</pre>

However, need for now also to set slurm parameters inside the script, or within the interactive job:
<pre>
export SLURM_TASKS_PER_NODE=48
export SLURM_CPUS_ON_NODE=48
</pre>

==== Script for 1 GPU ====
<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=gpu_job # Job name
#SBATCH --account=my_account # Account name
#SBATCH --partition=gpu-general # Partition name
#SBATCH --time=02:00:00 # Max run time
#SBATCH --ntasks=1 # Number of tasks
#SBATCH --nodes=1 # Number of nodes
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --gres=gpu:1 # Number of GPUs
#SBATCH --mem-per-cpu=4G # Memory per CPU
#SBATCH --output=my_job_%j.out # Output file
#SBATCH --error=my_job_%j.err # Error file

module load python/python-3.8

echo "Starting GPU job"
echo "Job ID: $SLURM_JOB_ID"
echo "Running on nodes: $SLURM_JOB_NODELIST"
echo "Allocated CPUs: $SLURM_JOB_CPUS_PER_NODE"

# Your GPU commands go here

echo "Job completed"
</syntaxhighlight>
For excluding a node, one may add the following
<syntaxhighlight>
#SBATCH --exclude=compute-0-[100-103],compute-0-67
</syntaxhighlight>

===Importance of Correct RAM Usage in Jobs===

When writing SLURM job scripts, it's crucial to understand and correctly specify the memory requirements for your job.

Proper memory allocation ensures efficient resource usage and prevents job failures due to out-of-memory (OOM) errors.

==== Why Correct RAM Usage Matters ====

* '''Resource Efficiency''': Allocating the right amount of memory helps in optimal resource utilization, allowing more jobs to run simultaneously on the cluster.
* '''Job Stability''': Underestimating memory requirements can lead to OOM errors, causing your job to fail and waste computational resources.
* '''Performance''': Overestimating memory needs can lead to underutilization of resources, potentially delaying other jobs in the queue.

==== How to Specify Memory in SLURM ====

* '''--mem''': Specifies the total memory required for the job.
* '''--mem-per-cpu''': Specifies the memory required per CPU.

'''Example''':<syntaxhighlight lang="bash">
#SBATCH --mem=4G # Total memory for the job
#SBATCH --mem-per-cpu=2G # Memory per CPU
</syntaxhighlight>

===Interactive Jobs===
<syntaxhighlight lang="bash">
#Start an interactive session:
srun --ntasks=1 -p power-general -A power-general-users --pty bash

#Specify a compute node:
srun --ntasks=1 -p power-general -A power-general-users --nodelist="compute-0-12" --pty bash

#Using GUI:
srun --ntasks=1 -p power-general -A power-general-users --x11 /bin/bash

</syntaxhighlight>

=== Submitting RELION Jobs===

To submit a RELION job interactively on the <code>gpu-relion</code> queue with X11 forwarding, use the following steps:<syntaxhighlight lang="bash">
#Start an interactive session with X11:
srun --ntasks=1 -p gpu-relion -A your_account --x11 --pty bash
#Load the RELION module:
module load relion/relion-4.0.1
#Launch RELION:
relion

</syntaxhighlight>

==Running matlab example==
In this example there are 3 files:

myTable.m ⇒ This matlab file calculates something
<pre>
fprintf('=======================================\n');
fprintf(' a b c d \n');
fprintf('=======================================\n');
while 1
for j = 1:10
a = sin(10*j);
b = a*cos(10*j);
c = a + b;
d = a - b;
fprintf('%+6.5f %+6.5f %+6.5f %+6.5f \n',a,b,c,d);
end
end
fprintf('=======================================\n');
</pre>

my_table_script.sh ⇒ This script executes the matlab program. Need just to run qsub with this script
<pre>
#!/bin/bash

#SBATCH --mem=50mg
#SBATCH --partition powers-general
#SBATCH -A power-general-users
hostname

cd /a/home/cc/tree/taucc/staff/dvory/matlab

matlab -nodisplay -nosplash -nodesktop -r "run(myTable());exit;"
</pre>

run_in_loop.sh ⇒ However, one may also generate many jobs with this file
<pre>
#!/bin/bash

for i in {1..100}

do

sbatch my_table_script.sh

done
</pre>

Running my job is with the command (after doing chmod +x 'run_in_loop.sh'):
<pre>
./run_in_loop.sh
</pre>

==AlphaFold==

AlphaFold is a deep learning tool designed for predicting protein structures.

'''Guides:'''

[https://hpcguide.tau.ac.il/index.php?title=Alphafold AlphaFold Guide]

[https://hpcguide.tau.ac.il/index.php?title=Alphafold3 AlphaFold3 Guide]

==Common SLURM Commands==
<syntaxhighlight lang="bash">
#View all queues (partitions):
sinfo
#View all jobs:
squeue
#View details of a specific job:
scontrol show job <job_number>
#Get information about partitions:
scontrol show partition
</syntaxhighlight>

== Troubleshooting & Tips ==

=== Common Errors ===

# <code>srun: error: Unable to allocate resources: No partition specified or system default partition</code> '''Solution:''' Always specify a partition. Example: <code>srun --pty -c 1 --mem=2G -p power-general /bin/bash</code>
# Job failed, and upon doing scontrol show job job_id or when running sacct -j job_id -o JobID,JobName,State%20 you see: <code>JobState=OUT_OF_MEMORY Reason=OutOfMemory</code> or :<syntaxhighlight lang="bash">
JobID JobName State
------------ ---------- --------------------
71 oom_test OUT_OF_MEMORY
71.batch batch OUT_OF_MEMORY
71.extern extern COMPLETED

</syntaxhighlight>it means that the ram requested for the job was not enough, please resubmit the job again with more ram. see [https://wikihpc.tau.ac.il/index.php?title=Slurm_user_guide#Estimating_RAM_Usage below] for help with understanding how much ram your job may need.

=== Chain Jobs ===
Use the <code>--depend</code> flag to set job dependencies.

'''Example:'''
<syntaxhighlight lang="bash">
sbatch --ntasks=1 --time=60 -p power-general -A power-general-users --depend=45001 do_work.bash
</syntaxhighlight>

=== Always Specify Resources ===
When submitting jobs, ensure you include all required resources like partition, memory, and CPUs to avoid job failures.

=== Attaching to Running Jobs ===
If you need to monitor or interact with a running job, use <code>sattach</code>. This command allows you to attach to a job's input, output, and error streams in real-time.

'''Example:'''
<syntaxhighlight lang="bash">
sattach <job_id>
</syntaxhighlight>

To view job steps of a specific job, use the following command:

<syntaxhighlight lang="bash">
scontrol show job <job_id>
</syntaxhighlight>

Look for sections labeled "StepId" within the output.

'''For specific job steps, use:'''
<syntaxhighlight lang="bash">
sattach <job_id.step_id>
</syntaxhighlight>

'''Note:''' <code>sattach</code> is particularly useful for interactive jobs, where you can provide input directly. For non-interactive jobs, it acts like <code>tail -f</code>, allowing you to monitor the output stream.

=== Estimating RAM Usage ===

When writing SLURM job scripts, it's crucial to understand and correctly specify the memory requirements for your job. Proper memory allocation ensures efficient resource usage and prevents job failures due to out-of-memory (OOM) errors.

==== Tips for Estimating RAM Usage ====

* Check Application Documentation: Refer to the official documentation or user guides for memory-related information.
* Run a Small Test Job: Submit a smaller version of your job and monitor its memory usage using commands like `free -m`, `top`, or `htop`.
* Use Profiling Tools: Tools like `valgrind`, `gprof`, or built-in profilers can help you understand memory usage.
* Analyze Previous Jobs: Review SLURM logs and job statistics for insights into memory consumption of past jobs.
* Consult with Peers or Experts: Ask colleagues or experts who have experience with similar workloads.

==== Example: Monitoring Memory Usage ====
<syntaxhighlight lang="bash">
#!/bin/bash

#SBATCH --job-name=memory_test
#SBATCH --account=your_account
#SBATCH --partition=your_partition
#SBATCH --time=01:00:00
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=1
#SBATCH --mem=4G
#SBATCH --output=memory_test.out
#SBATCH --error=memory_test.err

# Monitor memory usage
echo "Memory usage before running the job:"
free -m

# Your application commands go here
# ./your_application

# Monitor memory usage after running the job
echo "Memory usage after running the job:"
free -m
</syntaxhighlight>

==== General Tips ====

* Start Small: Begin with a conservative memory request and increase it based on observed usage.
* Consider Peak Usage: Plan for peak memory usage to avoid OOM errors.
* Use SLURM's Memory Reporting: Use `sacct` to view memory usage statistics.

'''Example:'''
<syntaxhighlight lang="bash">
sacct -j <job_id> --format=JobID,JobName,MaxRSS,Elapsed
</syntaxhighlight>

Submitting vscode job on slurm

2025-04-03T12:31:38Z

Dvory: Created page with "2 options hereby presented to use vscode 1. vscode via using an interactive job, and loading vscode module 2. vscode via ssh proxyjump =vscode module= This method may be to..."

2 options hereby presented to use vscode

1. vscode via using an interactive job, and loading vscode module

2. vscode via ssh proxyjump

=vscode module=
This method may be too slow for the users, as noted by user luxembourg, however possible.

User needs to:

Login to powerslurm with '-X'
Request for interactive job with GUI,
<pre>
srun --partition power-general --x11 --pty bash
</pre>
Then on the compute node need to load module
<pre>
module load vscode/vscode-1.98.2
</pre>

Also need to define a cache dir, it may be done in one of the scratches, e.g.
<pre>
export XDG_RUNTIME_DIR=/scratch200/dvory
</pre>

So within the interactive job one may activate vscode by typing
<pre>
code
</pre>

=vscode via ssh proxyjump=

Both '''vscode''' and '''cursor''' behave similarly, and this approach is supposed to be good for both of them.

With this setup, a windows machine will be able to activate vscode/cursor in a compute node in the cluster, having the same slurm environment variables as defined in a batch script.

* Every user may configure this, no need for root

==In powerslurm-login node==
===One time setup===
Install dropbear in userspace.

'''dropbear''' is used to launch a private SSH server inside the job
<pre>
cd ~
mkdir -p dropbear/install

# Build it from source
curl -LO https://matt.ucc.asn.au/dropbear/releases/dropbear-2022.83.tar.bz2
tar -xjf dropbear-2022.83.tar.bz2
cd dropbear-2022.83
./configure --prefix=$HOME/dropbear/install
make && make install

# Generate host key
$HOME/dropbear/install/bin/dropbearkey -t rsa -f $HOME/dropbear/install/server-key
</pre>

===Build a script===
Name of the script may be for example vscode_slurm_job.sh
<pre>
#!/bin/bash
#SBATCH --job-name=vscode
#SBATCH --partition=power-general
#SBATCH --cpus-per-task=8
#SBATCH --mem=64G
#SBATCH --output=slurm_vscode.log

DROPBEAR="$HOME/dropbear/install"
SLURM_ENV_FILE="$HOME/.slurm-env.bash"

# 🧠 Derive unique port from SLURM_JOB_ID
PORT=$(( 40000 + SLURM_JOB_ID % 10000 ))

# Save SLURM-related env vars to a file
env | awk -F= '$1~/^(SLURM|CUDA|NVIDIA_)/ { print "export " $0 }' > "$SLURM_ENV_FILE"

# Write VS Code shell script
cat <<EOF > "$HOME/.vscode-shell"
#!/bin/bash
source $SLURM_ENV_FILE
export XDG_RUNTIME_DIR=/tmp/\$USER-vscode-runtime
mkdir -p \$XDG_RUNTIME_DIR
chmod 700 \$XDG_RUNTIME_DIR
exec bash
EOF

chmod +x "$HOME/.vscode-shell"

# 📝 Write SSH connection instructions to a file
HOSTNAME=$(hostname -f)
echo "VS Code SSH setup info for this job:" > $HOME/vscode-connection-info.txt
echo "--------------------------------------" >> $HOME/vscode-connection-info.txt
echo "HostName: $HOSTNAME" >> $HOME/vscode-connection-info.txt
echo "Port: $PORT" >> $HOME/vscode-connection-info.txt
echo "Username: $USER" >> $HOME/vscode-connection-info.txt
echo "RemoteCommand: ~/.vscode-shell" >> $HOME/vscode-connection-info.txt
echo "ProxyJump: $USER@powerslurm-login.tau.ac.il" >> $HOME/vscode-connection-info.txt
echo "" >> $HOME/vscode-connection-info.txt
echo "Suggested ~/.ssh/config entry:" >> $HOME/vscode-connection-info.txt
cat <<EOF >> $HOME/vscode-connection-info.txt

Host slurm-vscode
HostName $HOSTNAME
User $USER
Port $PORT
ProxyJump $USER@powerslurm-login.tau.ac.il
RemoteCommand ~/.vscode-shell
IdentityFile ~/.ssh/your_key
StrictHostKeyChecking no
UserKnownHostsFile /dev/null
EOF

# Cleanup on exit
cleanup() {
rm -f "$SLURM_ENV_FILE"
}
trap 'cleanup' SIGTERM SIGINT EXIT

# Launch dropbear
$DROPBEAR/sbin/dropbear -r $DROPBEAR/server-key -F -E -w -s -p $PORT
</pre>

==In windows==
===Generate an SSH key===
Open PowerShell or Git Bash on Windows, and run:
<pre>
ssh-keygen -t rsa -b 4096
</pre>
Accept the default path (C:\Users\YourName\.ssh\id_rsa)

Choose a passphrase or leave empty

You now have:

Private key: C:\Users\YourName\.ssh\id_rsa ← Keep this secure!

Public key: C:\Users\YourName\.ssh\id_rsa.pub

Copy the content of id_rsa.pub to powerslurm-login
<pre>
vim ~/.ssh/authorized_keys
# Paste the key and save
</pre>
This key may also be copied to the compute node that is allocated for the job

===Create ~\.ssh\config file===
Set up ~/.ssh/config:

Create or edit C:\Users\<YourUsername>\.ssh\config, containing:
<pre>
Host slurm-vscode
HostName compute-0-xxx # Replace with the compute node
User dvory # Replace with actual username
ProxyJump dvory@powerslurm-login.tau.ac.il
Port 64321
IdentityFile ~/.ssh/id_rsa # Or correct private key
StrictHostKeyChecking no
UserKnownHostsFile /dev/null
</pre>

===Set permissions of .ssh/config===
Need to open cmd, and then run:
<pre>
icacls "%USERPROFILE%\.ssh\id_rsa" /reset
icacls "%USERPROFILE%\.ssh\id_rsa" /inheritance:r
icacls "%USERPROFILE%\.ssh\id_rsa" /grant:r "%USERNAME%:F"
icacls "%USERPROFILE%\.ssh\config" /reset
icacls "%USERPROFILE%\.ssh\config" /inheritance:r
icacls "%USERPROFILE%\.ssh\config" /grant:r "%USERNAME%:F"
</pre>

Download '''VS Code''' (https://code.visualstudio.com/) or '''Cursor''' (https://www.cursor.com/)

Open the application and install there, or make sure that thie application Remote-SSH is installed

===Installing Remote-SSH extension===
Launch your VS Code / Cursor app.
🔍 Open the Extensions Sidebar

Click the square icon on the left sidebar (or press Ctrl+Shift+X)

In the Search box, type:

Remote - SSH

📦 Install the Extension

Find: "Remote - SSH by Microsoft"

Click Install

Remote-SSH Extension in Marketplace

You’ll see a green install bar. Once done, you’ll see a new green "><" icon in the bottom-left — that’s the remote connection control.

===Test connection===
In VS Code or Cursor:

Hit F1 → “Remote-SSH: Connect to Host”

Select '''slurm-vscode'''
You should get connected to the compute node inside the running Slurm job.

==Running vscode==
===On powerslurm-login===
Submit the script:
<pre>
sbatch vscode_slurm_job.sh
</pre>
Check that the job is running, and find the allocated node:
<pre>
squeue --me
</pre>
The script that was submitted to run is supposed to generate a file named ~/vscode-connection-info.txt, containing all information needed for the proxyjump procedure.

Check compute node and port when the job is running, by running:
<pre>
cat ~/vscode-connection-info.txt
</pre>

You will see an output like:
<pre>
VS Code SSH setup info for this job:
--------------------------------------
HostName: compute-0-499.power5
Port: 43104
Username: dvory
RemoteCommand: ~/.vscode-shell
ProxyJump: dvory@powerslurm-login.tau.ac.il

Suggested ~/.ssh/config entry:

Host slurm-vscode
HostName compute-0-499.power5
User dvory
Port 43104
ProxyJump dvory@powerslurm-login.tau.ac.il
RemoteCommand ~/.vscode-shell
IdentityFile ~/.ssh/your_key
StrictHostKeyChecking no
UserKnownHostsFile /dev/null
</pre>

And would need to change accordingly the ~/.ssh/config file on windows

Let us assume it is running on node compute-0-499, port 43104 as was returned in the above example
===On windows===
Change file ~/.ssh/config according to the actual compute node that was allocated to the job
<pre>
Host slurm-vscode
HostName compute-0-499 # Replace with the compute node
User dvory # Replace with actual username
ProxyJump dvory@powerslurm-login.tau.ac.il
Port 43104 # Replace with actual port
IdentityFile ~/.ssh/id_rsa # Or correct private key
StrictHostKeyChecking no
UserKnownHostsFile /dev/null
</pre>

In vscode ot cursor

Press F1 → type and select:
<pre>
Remote-SSH: Connect to Host...
</pre>
Or click the green "><" icon (bottom-left)

Choose a host defined in your ~/.ssh/config (e.g., slurm-vscode)

Main Page

2025-04-03T12:30:14Z

Dvory:

'''Welcome to HPC Guide.'''

[[Linux basic commands]]

[[Public queues]]

[[Submitting a job to a queue]]

[[Submitting a job to a slurm queue]]

[[PBS-To-SLURM]]

[[Creaing and using conda environment]]

[[Palo Alto VPN for linux]]

[[Alphafold]]

[[Alphafold3]]

[[Using GPU]]

[[security installations]]

[[Install matlab on work station per matlab user]]

[[Submitting vscode job on slurm]]

[[Storage and scratch]]

This HPC Tutorial is designed for researchers at TAU who are in need of computational power (computer resources) and wish to explore and use our High Performance Computing (HPC) core facilities.
 
The audience may be completely unaware of the HPC concepts but must have some basic understanding of computers and computer programming.

'''What is HPC?'''

“High Performance Computing” (HPC) is computing on a “Supercomputer”,
 
a computer at the front line of contemporary processing capacity – particularly speed of calculation and available memory.
 
A computer cluster consists of a set of loosely or tightly connected computers that work together so that in many respects they can be viewed as a single system.
 
The components of a cluster are usually connected to each other through fast local area networks(“LAN”) with each node (computer used as a server) running its own instance of an operating system.
 
Computer clusters emerged as a result of convergence of a number of computing trends including the availability of low cost microprocessors,
 
high-speed networks, and software for high performance distributed computing.
 
Compute clusters are usually deployed to improve performance and availability over that of a single computer, while typically being more cost-effective than single computers of comparable speed or availability.

2024-12-26T09:16:34Z

Dvory: /* Basic Script */

== Accessing the System ==

To submit jobs to SLURM at Tel Aviv University, you need to access the system through one of the following login nodes:

* powerslurm-login.tau.ac.il
* powerslurm-login2.tau.ac.il

=== Requirements for Access ===

* '''Group Membership''': You must be part of the "power" group to access the resources.
* '''University Credentials''': Use your Tel Aviv University username and password to log in.

These login nodes are your starting point for submitting jobs, checking job status, and managing your SLURM tasks.

=== SSH Example ===

To access the system using SSH, use the following example:

<syntaxhighlight lang="bash">
# Replace 'your_username' with your actual Tel Aviv University username
ssh your_username@powerslurm-login.tau.ac.il
</syntaxhighlight>

If you want to connect to the second login node, use:

<syntaxhighlight lang="bash">
# Replace 'your_username' with your actual Tel Aviv University username
ssh your_username@powerslurm-login2.tau.ac.il
</syntaxhighlight>

If you have an SSH key set up for password-less login, you can specify it like this:

<syntaxhighlight lang="bash">
# Replace 'your_username' and '/path/to/your/private_key' accordingly
ssh -i /path/to/your/private_key your_username@powerslurm-login.tau.ac.il
</syntaxhighlight>

== Environment Modules ==

Environment Modules in SLURM allow users to dynamically modify their shell environment, providing an easy way to load and unload different software applications, libraries, and their dependencies. This system helps avoid conflicts between software versions and ensures the correct environment for running specific applications.

Here are some common commands to work with environment modules:<syntaxhighlight lang="bash">
#List Available Modules: To see all the modules available on the system, use:
module avail

#To search for a specific module by name (e.g., `gcc`), use:
module avail gcc/gcc-12.1.0

#Get Detailed Information About a Module: The `module spider` command provides detailed information about a module, including versions, dependencies, and descriptions:
module spider gcc/gcc-12.1.0

#View Module Settings: To see what environment variables and settings will be modified by a module, use:
module show gcc/gcc-12.1.0

#Load a Module: To set up the environment for a specific software, use the `module load` command. For example, to load GCC version 12.1.0:
module load gcc/gcc-12.1.0

#List Loaded Modules: To view all currently loaded modules in your session, use:
module list

#Unload a Module: To unload a specific module from your environment, use:
module unload gcc/gcc-12.1.0

#Unload All Modules:** If you need to clear your environment of all loaded modules, use:
module purge

</syntaxhighlight>By using these commands, you can easily manage the software environments needed for different tasks, ensuring compatibility and reducing potential conflicts between software versions.

== Basic Job Submission Commands ==

=== Finding Your Account and Partition ===

Before submitting a job, you need to know which partitions you have permission to use.

Run the command `<code>check_my_partitions</code>` to view a list of all the partitions you have permission to send jobs to.

== Submitting Jobs==
sbatch: Submits a job script for batch processing.

'''Example''':<syntaxhighlight lang="bash">

sbatch --ntasks=1 --time=10 -p power-general -A power-general-users pre_process.bash
# This command submits pre_process.bash to the power-general partition for 10 minutes.
# With 1 GPU:
sbatch --gres=gpu:1 -p gpu-general -A gpu-general-users gpu_job.sh

</syntaxhighlight>

===Writing SLURM Job Scripts===

Here is a simple job script example:

==== Basic Script====
<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=my_job # Job name
#SBATCH --account=power-general-users # Account name
#SBATCH --partition=power-general # Partition name
#SBATCH --time=02:00:00 # Max run time (hh:mm:ss)
#SBATCH --ntasks=1 # Number of tasks
#SBATCH --nodes=1 # Number of nodes
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --mem-per-cpu=4G # Memory per CPU
#SBATCH --output=my_job_%j.out # Output file
#SBATCH --error=my_job_%j.err # Error file

echo "Starting my SLURM job"
echo "Job ID: $SLURM_JOB_ID"
echo "Running on nodes: $SLURM_JOB_NODELIST"
echo "Allocated CPUs: $SLURM_JOB_CPUS_PER_NODE"

# Your application commands go here
# ./my_program

echo "Job completed"
</syntaxhighlight>

==== Script for 1 GPU ====
<syntaxhighlight lang="bash">
#!/bin/bash
#SBATCH --job-name=gpu_job # Job name
#SBATCH --account=my_account # Account name
#SBATCH --partition=gpu-general # Partition name
#SBATCH --time=02:00:00 # Max run time
#SBATCH --ntasks=1 # Number of tasks
#SBATCH --cpus-per-task=1 # CPUs per task
#SBATCH --gres=gpu:1 # Number of GPUs
#SBATCH --mem-per-cpu=4G # Memory per CPU
#SBATCH --output=my_job_%j.out # Output file
#SBATCH --error=my_job_%j.err # Error file

module load python/python-3.8

echo "Starting GPU job"
echo "Job ID: $SLURM_JOB_ID"
echo "Running on nodes: $SLURM_JOB_NODELIST"
echo "Allocated CPUs: $SLURM_JOB_CPUS_PER_NODE"

# Your GPU commands go here

echo "Job completed"
</syntaxhighlight>

===Importance of Correct RAM Usage in Jobs===

When writing SLURM job scripts, it's crucial to understand and correctly specify the memory requirements for your job.

Proper memory allocation ensures efficient resource usage and prevents job failures due to out-of-memory (OOM) errors.

==== Why Correct RAM Usage Matters ====

* '''Resource Efficiency''': Allocating the right amount of memory helps in optimal resource utilization, allowing more jobs to run simultaneously on the cluster.
* '''Job Stability''': Underestimating memory requirements can lead to OOM errors, causing your job to fail and waste computational resources.
* '''Performance''': Overestimating memory needs can lead to underutilization of resources, potentially delaying other jobs in the queue.

==== How to Specify Memory in SLURM ====

* '''--mem''': Specifies the total memory required for the job.
* '''--mem-per-cpu''': Specifies the memory required per CPU.

'''Example''':<syntaxhighlight lang="bash">
#SBATCH --mem=4G # Total memory for the job
#SBATCH --mem-per-cpu=2G # Memory per CPU
</syntaxhighlight>

===Interactive Jobs===
<syntaxhighlight lang="bash">
#Start an interactive session:
srun --ntasks=1 -p power-general -A power-general-users --pty bash

#Specify a compute node:
srun --ntasks=1 -p power-general -A power-general-users --nodelist="compute-0-12" --pty bash

#Using GUI:
srun --ntasks=1 -p power-general -A power-general-users --x11 /bin/bash

</syntaxhighlight>

=== Submitting RELION Jobs===

To submit a RELION job interactively on the <code>gpu-relion</code> queue with X11 forwarding, use the following steps:<syntaxhighlight lang="bash">
#Start an interactive session with X11:
srun --ntasks=1 -p gpu-relion -A your_account --x11 --pty bash
#Load the RELION module:
module load relion/relion-4.0.1
#Launch RELION:
relion

</syntaxhighlight>

==AlphaFold==

AlphaFold is a deep learning tool designed for predicting protein structures.

'''Guide:'''
[https://hpcguide.tau.ac.il/index.php?title=Alphafold AlphaFold Guide]

==Common SLURM Commands==
<syntaxhighlight lang="bash">
#View all queues (partitions):
sinfo
#View all jobs:
squeue
#View details of a specific job:
scontrol show job <job_number>
#Get information about partitions:
scontrol show partition
</syntaxhighlight>

== Troubleshooting & Tips ==

=== Common Errors ===

# <code>srun: error: Unable to allocate resources: No partition specified or system default partition</code> '''Solution:''' Always specify a partition. Example: <code>srun --pty -c 1 --mem=2G -p power-general /bin/bash</code>
# Job failed, and upon doing scontrol show job job_id or when running sacct -j job_id -o JobID,JobName,State%20 you see: <code>JobState=OUT_OF_MEMORY Reason=OutOfMemory</code> or :<syntaxhighlight lang="bash">
JobID JobName State
------------ ---------- --------------------
71 oom_test OUT_OF_MEMORY
71.batch batch OUT_OF_MEMORY
71.extern extern COMPLETED

</syntaxhighlight>it means that the ram requested for the job was not enough, please resubmit the job again with more ram. see [https://wikihpc.tau.ac.il/index.php?title=Slurm_user_guide#Estimating_RAM_Usage below] for help with understanding how much ram your job may need.

=== Chain Jobs ===
Use the <code>--depend</code> flag to set job dependencies.

'''Example:'''
<syntaxhighlight lang="bash">
sbatch --ntasks=1 --time=60 -p power-general -A power-general-users --depend=45001 do_work.bash
</syntaxhighlight>

=== Always Specify Resources ===
When submitting jobs, ensure you include all required resources like partition, memory, and CPUs to avoid job failures.

=== Attaching to Running Jobs ===
If you need to monitor or interact with a running job, use <code>sattach</code>. This command allows you to attach to a job's input, output, and error streams in real-time.

'''Example:'''
<syntaxhighlight lang="bash">
sattach <job_id>
</syntaxhighlight>

To view job steps of a specific job, use the following command:

<syntaxhighlight lang="bash">
scontrol show job <job_id>
</syntaxhighlight>

Look for sections labeled "StepId" within the output.

'''For specific job steps, use:'''
<syntaxhighlight lang="bash">
sattach <job_id.step_id>
</syntaxhighlight>

'''Note:''' <code>sattach</code> is particularly useful for interactive jobs, where you can provide input directly. For non-interactive jobs, it acts like <code>tail -f</code>, allowing you to monitor the output stream.

=== Estimating RAM Usage ===

When writing SLURM job scripts, it's crucial to understand and correctly specify the memory requirements for your job. Proper memory allocation ensures efficient resource usage and prevents job failures due to out-of-memory (OOM) errors.

==== Tips for Estimating RAM Usage ====

* Check Application Documentation: Refer to the official documentation or user guides for memory-related information.
* Run a Small Test Job: Submit a smaller version of your job and monitor its memory usage using commands like `free -m`, `top`, or `htop`.
* Use Profiling Tools: Tools like `valgrind`, `gprof`, or built-in profilers can help you understand memory usage.
* Analyze Previous Jobs: Review SLURM logs and job statistics for insights into memory consumption of past jobs.
* Consult with Peers or Experts: Ask colleagues or experts who have experience with similar workloads.

==== Example: Monitoring Memory Usage ====
<syntaxhighlight lang="bash">
#!/bin/bash

#SBATCH --job-name=memory_test
#SBATCH --account=your_account
#SBATCH --partition=your_partition
#SBATCH --time=01:00:00
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=1
#SBATCH --mem=4G
#SBATCH --output=memory_test.out
#SBATCH --error=memory_test.err

# Monitor memory usage
echo "Memory usage before running the job:"
free -m

# Your application commands go here
# ./your_application

# Monitor memory usage after running the job
echo "Memory usage after running the job:"
free -m
</syntaxhighlight>

==== General Tips ====

* Start Small: Begin with a conservative memory request and increase it based on observed usage.
* Consider Peak Usage: Plan for peak memory usage to avoid OOM errors.
* Use SLURM's Memory Reporting: Use `sacct` to view memory usage statistics.

'''Example:'''
<syntaxhighlight lang="bash">
sacct -j <job_id> --format=JobID,JobName,MaxRSS,Elapsed
</syntaxhighlight>

Submitting a job to a slurm queue

2024-09-22T13:49:07Z

Dvory: /* Basic Job Submission Commands */

2024-04-15T13:16:59Z

Dvory: Created page with "Besides each user's home directory, one may purchase disk storage, prices are according to https://view.monday.com/4073193937-33252df4e02cadb641ff891627342c96?r=use1 netapp st..."

Besides each user's home directory, one may purchase disk storage, prices are according to https://view.monday.com/4073193937-33252df4e02cadb641ff891627342c96?r=use1
netapp storage usually contains snapshots (described in https://view.monday.com/4073193937-33252df4e02cadb641ff891627342c96?r=use1)
The storage may also be backed up by legato server, as described in https://computing.tau.ac.il/infrastructure_backup

Apart from the storage there are /scratch partitions:
/scratch100
/scratch200
/scratch300
Which may be used as scratch, i.e. for temporary usage.
scratches are not being backed up.

In some work stations and compute nodes there is also a local /scratch folder, which is used in a similar way.