Cluster Documentation 2024: Difference between revisions

From ITTC Help
Jump to navigation Jump to search
(Redirected page to Cluster Documentation)
Tag: New redirect
 
Line 1: Line 1:
== Introduction ==
#REDIRECT [[Cluster Documentation]]
 
The I2S Research Cluster is located in the ACF, and provides HPC resources to members of the center. The cluster uses the Slurm workload manager, which is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for Linux clusters. The cluster is composed of a variety of hardware types, with core counts ranging from 16 to 32 cores per node. In addition, there is specialized hardware including Nvidia graphics cards for GPU computing, Infiniband for low latency/high throughput parallel computing, and large memory systems with up to 512 GB of RAM.
 
== Getting Help ==
You must have an ITTC/I2S Account and be part of the cluster group. If you have any questions about the I2S Research Cluster, feel free to email [mailto:clusterhelp@ittc.ku.edu clusterhelp@ittc.ku.edu] for assistance.
 
== Connecting ==
=== SSH ===
This connection method leaves you with a terminal window accepting only keyboard input in a command line interface. This is the preferred way to use the cluster.
 
From a Linux or Mac, you can simply run this from a terminal window:
<pre>
ssh <your i2s username here>@front1.ittc.ku.edu
</pre>
 
In Windows, you can either use Putty or enter the command above into Windows Terminal.
 
The above example uses the front1 front end, but we'd recommend users to use the front2 front end as well to balance out usage.
 
You'll be left with a shell on the front-end node. Do not run your computationally-expensive tasks directly on these nodes, since they are shared with all of the cluster users. Instead, create an allocation on a compute node, detailed in the Job Submission Guide.
 
=== Available front-end node hostnames ===
<pre>
front1.ittc.ku.edu
front2.ittc.ku.edu
</pre>
 
=== Campus ===
 
If you are connecting from any of the University of Kansas campuses, you may connect to either host directly.
 
=== Off-Campus ===
 
If you wish to connect to any of the I2S clusters from off campus, you must connect to KU Anywhere (VPN) before attempting to connect to the cluster. More information is available at [https://technology.ku.edu/catalog/ku-anywhere-vpn KU Anywhere].
 
== Job Submission Guide ==
=== PBS/Torque and Slurm ===
A translation for common PBS/Torque commands to Slurm commands can be found [https://slurm.schedmd.com/rosetta.pdf here]. This provides a quick guide for those who are familiar with PBS/Torque, but new to the Slurm scheduler.
 
 
 
=== Submitting Jobs ===
 
To submit jobs to the cluster, you can either write a script and submit it using sbatch:
<pre>
[username@front1 ~]$ sbatch script.sh
</pre>
 
Or, you can submit jobs interactively from the command line using srun:
<pre>
[username@front1 ~]$ srun echo Hello World!
</pre>
 
Job scripts use parameters (denoted by <code>#SBATCH</code>) in the script file to requested job resources, while interactive jobs request resources with command line parameters. When no resources are requested, a default set is automatically allocated for the job.
 
<div id="DefaultResources">
;This default resource set includes <b>:</b>
*The job's name is set the same as the script file name, or, if the job was started with srun, then the job name is the same as the first command (in the case of the example above, the name would be set to 'echo').<br/>
*The job is scheduled in the default intel queue.<br/>
*The job is allocated 1 core on 1 node with 2GB of memory.<br/>
*The job is allocated 1 day to run.<br/>
*The job redirects stdout and stderr to the same output file if the job is submitted with sbatch. If srun is used, then both will be printed to the screen.<br/>
*The job's output file name takes the form "slurm-<i>jobid</i>.out", and is created in the same directory as the job script.<br/>
</div>
 
==== <h3>srun</h3> ====
 
<code>srun</code> can be used to run any single task on a cluster node, but it is most useful for launching interactive GUI or bash sessions. Here is an srun example run on front1:
 
<pre>
[username@front1 ~]$ srun -p intel -N 1 -n 1 -c 4 --mem 4G --pty /bin/bash
[username@n097 ~]$
</pre>
 
;The options used in this example are all detailed below<b>:</b>:
:; -p        : Specifies a partition, or queue to create the job in. The current cluster partitions available are <code>intel</code>, <code>amd</code>, <code>bigm</code>, and <code>gpu</code>. For more information on the cluster queues, see the [[#Cluster Partitions|partitions section]] below.
:; -N        : This sets the number of requested nodes for the interactive session.
:; -n        : Specifies the number of tasks or processes to run on each allocated node.
:; -c        : Sets the number of requested cpus per task
:; --mem    : This specifies the requested memory per node. Memory amounts can be given in Kilobytes (K), Megabytes (M), and Gigabytes (G).
:; --pty    : This option puts the srun session in pseudo-terminal mode. It is recommended to use this option if you are running an interactive shell session.
:; /bin/bash : The last option in an srun invocation is the program that srun will execute on the requested node. In this case, bash is specified to start an interactive shell session.
 
srun is used to submit both interactive and non-interactive jobs. When it is run directly on the command line as shown above, an interactive session is started on a cluster node. When it is used in a job submission script, it starts a non-interactive session.
 
==== <h3>sbatch</h3> ====
 
<code>sbatch</code> is used to submit jobs to the cluster using a script file. Below is an example job submission script:
 
<pre>
#!/bin/bash
#SBATCH -p intel
#SBATCH -N 1
#SBATCH -n 1
#SBATCH -c 1
#SBATCH --mem=1GB
#SBATCH -t 00:20:00
#SBATCH -J test_job
#SBATCH -o slurm-%j.out
echo "Job ${SLURM_JOB_ID} ran on ${HOSTNAME}"
</pre>
 
Example output:
 
<pre>
[username@front1 ~]$ sbatch test_job.sh
[username@front1 ~]$ cat slurm-47491.out
Job 47491 ran on n097
[username@front1 ~]$
</pre>
 
This script requests one node with one core, and 1GB of memory. <code>-J</code> is used to specify the job name that appears in the job queue, while <code>-o</code> specifies the log file name for the job. <code>%j</code> in the job output file name is replaced with the Slurm job id when the scheduler processes the script. The variable <code>SLURM_JOB_ID</code> used in the example output is an environment variable set by the Slurm scheduler for each job.
 
To run this example script, copy its contents into a file in your home directory (test_job.sh for example). Log in to either front1.ittc.ku.edu or front2.ittc.ku.edu with your ITTC credentials, and run the command <code>sbatch test_job.sh</code>. The job output log will be saved in the same directory as the job submission script, and should contain similar output to the example above.
 
sbatch job scripts can run programs directly, as show above, but it is also possible to use srun within job submission scripts to run programs. Using srun in a job script allows for fine-grained resource control over parallel tasks run in a job script. An example is shown below:
 
<pre>
#!/bin/bash
#SBATCH -p intel
#SBATCH -N 1
#SBATCH -n 2
#SBATCH -c 1
#SBATCH --mem=2GB
#SBATCH -t 00:20:00
#SBATCH -J test_job
#SBATCH -o slurm-%j.out
 
srun -n 1 --mem=1G echo "Task 1 ran" &
srun -n 1 --mem=1G echo "Task 2 ran" &
 
wait
</pre>
 
When the sbatch script is submitted, both srun invocations will run at the same time, splitting the resources requested at the top of the script file. This method is useful for launching a small number of related jobs at once from the same script, but does not scale well with a large number of jobs. The Job Array section below goes into more depth on running large numbers of parallel jobs on the cluster.
 
When using srun within a job submission script, you need to specify what portion of the resources each srun invocation is allocated. If more resources are requested by srun than are made available by the <code>#SBATCH</code> parameters, then some jobs may wait to run, or attempt to share resources with already running jobs. In the example above, two tasks and 2GB of memory are requested. In the srun commands below the resource request, we specify how much memory and how many tasks are allocated to each job.
 
The sbatch options shown in these example scripts are just the tip on the iceberg in terms of what is available. For the full listing of sbatch parameters, see the official Slurm [https://slurm.schedmd.com/sbatch.html sbatch documentation]
 
;Here is a brief list of other common options that may be useful<b>:</b>:
:; -C                : Specifies a node constraint. This can be used to specify cpu architecture, and instruction set.
:; -D                : Specifies the path to the log file destination directory. This can be an absolute path, or a relative path from the job submission script directory.
:; --gres            : Used to request GPU resources. See [[#GPU Jobs|this example]] for more information on running GPU jobs.
:; --cores-per-socket : Sets the requested number of cores per cpu socket.
:; --mem-per-cpu      : This specifies the memory allocated to each cpu in the interactive session. It has the same memory specification syntax as <code>--mem</code>.
:; --mail-type        : Sets when the user is to be mailed job notifications. NONE, BEGIN, END, FAIL, REQUEUE, TIME_LIMIT, TIME_LIMIT_90, TIME_LIMIT_80, and TIME_LIMIT_50 are all valid options
:; --mail-user        : Specifies the user account to email when job notification emails are sent.
 
==== <h3>Job Arrays</h3> ====
Submitting a large number of cluster jobs at once has two general approaches. The first is to submit jobs to the scheduler using srun in a loop on the command line. The preferable, and more powerful approach uses job arrays to submit large blocks of jobs all at once with the sbatch command.
 
The <code>--array</code> parameter for sbatch allows the scheduler to queue up hundreds to thousands of jobs with the same resource requests. This method is much less taxing on the cluster scheduler, and simplifies the process of submitting a large number of jobs all at once. These arrays usually consist of the same program fed different parameters dictated by the job array indicies.
 
An example job array script is shown below:
 
<pre>
#!/bin/bash
#SBATCH -p intel
#SBATCH -N 1
#SBATCH -n 1
#SBATCH -c 1
#SBATCH --mem=1G
#SBATCH -t 00:20:00
#SBATCH -J test_job
#SBATCH -o logs/%A_%a.out
#SBATCH --array=1-4
 
echo Job ${SLURM_ARRAY_TASK_ID} used $(awk "NR == ${SLURM_ARRAY_TASK_ID} {print \$0}" ${SLURM_SUBMIT_DIR}/parameters)
</pre>
Example output:
<pre>
[username@front1 ~]$ sbatch array_test.sh
[username@front1 ~]$ cd logs/
[username@front1 logs]$ ls
49219_1.out  49219_2.out  49219_3.out  49219_4.out
[username@front1 logs]$ cat *
Job 1 used line 1 parameters
Job 2 used line 2 parameters
Job 3 used line 3 parameters
Job 4 used line 4 parameters
[username@front1 logs]$
</pre>
Parameters file:
<pre>
line 1 parameters
line 2 parameters
line 3 parameters
line 4 parameters
</pre>
 
In this example, the <code>%A</code> and <code>%a</code> symbols in the job log file path are replaced by the scheduler with the job array id and job array indicie respectively for each job in the array. The <code>--array</code> option specifies the creation of a job array which consists of four identical jobs with indices ranging from 1 to 4. Each job in the array is created with the same resource request at the top of the file, and runs the same bash command at the bottom of the script file. The <code>echo</code> command prints out the <code>SLURM_ARRAY_TASK_ID</code>  (or job array indicie) environment variable of each job, along with one line from a file called "parameters". The <code>awk</code> command within the echo selects the line in the parameters file with the line number that matches the job array indicie value. This technique can be used to feed in specific parameters to different jobs within a job array.
 
Another way of generating program parameters for job arrays is through arithmetic. For example, if you wanted to define a minimum and maximum value a job needed to loop through based on its indicie value, in your job script, you may include something like this:
 
<pre>
MAX=$(echo "${SLURM_ARRAY_TASK_ID} * 1000" | bc)
MIN=$(echo "$({SLURM_ARRAY_TASK_ID} - 1) * 1000" | bc)
 
for (( i=$MIN; i<$MAX; i++ )); do
  # Perform calculations...
done
</pre>
 
<div id="Cluster Partitions">
 
=== Cluster Partitions ===
Cluster partitions, or queues, are sets of nodes in the cluster grouped by their features. Currently, there are four partition in the ITTC cluster: intel, amd, bigm, and gpu. The intel and amd partitions are made up of nodes that contain exclusively intel and amd cpus respectively. The bigm queue is made up of nodes with RAM from 256 to 500GB, and the gpu partition contains nodes with Nvidia gpu co-processors. Partitions can be specified in a job script with the <code>-p</code> option:
<pre>
#SBATCH -p intel
</pre>
They can also be specified in interactive sessions:
<pre>
srun -p intel -N 1 -n 1 --pty /bin/bash
</pre>
Partitions allow for high-level constraints on job hardware, but lack fine-grained control over things like cpu and gpu architecture.
</div>
==== Job Constraints ====
Job constraints allow precise specification for what hardware a job should run on. Cpu architectures and instruction sets can be requested, as well as the networking type, node manufacturer, and memory. Specifying hardware constraints is done with the <code>-C</code> option:
<pre>
#SBATCH -C "intel"
</pre>
Multiple constraints can also be specified at once:
<pre>
srun -C "intel&ib" --pty /bin/bash
</pre>
In this example, the <code>&</code> symbol between the two constraints specifies that both should be fulfilled for the job to run. The <code>|</code> symbol can also be used to specify that either one or the other constraint can be fulfilled. Additionally, square-brackets can be used to group together constraints. Here is an example combining all three:
<pre>
#SBATCH -C "[intel&ib]|[amd&eth_10g]"
</pre>
 
<b>Available constraints:</b>
{| valign="top"
|-
|<ul valign="top" ><b>Instruction Set</b><ul><li>sse3<li>sse4_1<li>sse4_2<li>sse4a<li>avx</ul></ul>
|<ul><b>CPU Brand/Cores</b><ul><li>intel<li>amd<li>intel8<li>amd8<li>intel12<li>intel16<li>intel20</ul></ul>
|<ul><b>Networking</b><ul><li>ib<li>ib_ddr<li>ib_qdr<li>noib<li>eth_10g</ul></ul>
|<ul><b>Manufacturer/CPU Brand/Cores/Memory</b><ul><li>del_int_8_16<li>del_int_8_24<li>del_int_12_24<li>asu_int_12_32<li>sup_int_12_32<li>asu_int_12_128<li>del_int_16_64<li>del_int_16_256<li>del_int_20_256<li>del_int_16_512<li>del_int_20_128<li>del_amd_8_16</ul></ul>
|}
 
=== GPU Jobs ===
Instead of using hardware constraints, GPUs are specified with <b>G</b>eneric <b>Res</b>ource (<code>gres</code>) requests. Below is an example of an interactive GPU job request:
<pre>
srun -p gpu --gres="gpu:k20:2" --pty /bin/bash
</pre>
This request specifies two Nvidia K20 GPUs in the GPU queue for the interactive session, along with the [[#DefaultResources|default job resources]]. The <code>--gres</code> option allows the specification of a the GPU model and number through a colon-delimited list. Below is a job script example:
<pre>
#SBATCH -p gpu
#SBATCH --gres="gpu:k40:1"
</pre>
The GPU partition must be specified when requesting GPUs, otherwise the scheduler will reject the job. Whenever a job is started on a GPU node, the environment variable <code>CUDA_VISIBLE_DEVICES</code> is set to contain a comma-delimited list of the GPUs allocated to the current job. Information about these GPUs can be viewed by running <code>nvidia-smi</code>.
 
Here is example output from the <code>srun</code> example above:
<pre>
[username@front1 ~]$ srun -p gpu --gres="gpu:k20:2" --pty /bin/bash
[username@g002 ~]$ echo $CUDA_VISIBLE_DEVICES
1,2
[username@g002 ~]$ nvidia-smi
Fri Jan 20 16:23:01 2017     
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 367.48                Driver Version: 367.48                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|        Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|  0  Tesla K20m          Off  | 0000:02:00.0    Off |                    0 |
| N/A  30C    P0    47W / 225W |      0MiB /  4742MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|  1  Tesla K20m          Off  | 0000:03:00.0    Off |                    0 |
| N/A  29C    P0    47W / 225W |      0MiB /  4742MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|  2  Tesla K20m          Off  | 0000:83:00.0    Off |                    0 |
| N/A  28C    P0    48W / 225W |      0MiB /  4742MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|  3  Tesla K20m          Off  | 0000:84:00.0    Off |                    0 |
| N/A  28C    P0    51W / 225W |      0MiB /  4742MiB |    100%      Default |
+-------------------------------+----------------------+----------------------+
                                                                             
+-----------------------------------------------------------------------------+
| Processes:                                                      GPU Memory |
|  GPU      PID  Type  Process name                              Usage      |
|=============================================================================|
|  No running processes found                                                |
+-----------------------------------------------------------------------------+
[username@g002 ~]$
</pre>
 
Currently, there are seven different GPU models available in the cluster:
{|
|-
|<ul><b>Gpu Models:</b><ul><li>k20<li>k40<li>k80<li>titanxp<li>p100<li>v100s<li>titanrtx<li>a100</ul></ul>
|}
 
=== MPI Jobs ===
If you want to use multiple processors with MPI, you need to request multiple tasks with the -n option.  You will also need to load a version of MPI (ex: OpenMPI).  An example MPI job script is shown below:
 
<pre>
#!/bin/bash
 
#SBATCH -p intel
#SBATCH -n 4
#SBATCH --mem=1GB
#SBATCH -t 00:05:00
#SBATCH -J mpi_example
#SBATCH -o slurm-%j.out
 
module load OpenMPI
mpirun $HOME/helloworld
</pre>
 
The script above will launch four tasks of helloworld.  Below is the example output:
 
<pre>
Hello world from processor n097.local, rank 0 out of 4 processors
Hello world from processor n097.local, rank 1 out of 4 processors
Hello world from processor n097.local, rank 2 out of 4 processors
Hello world from processor n097.local, rank 3 out of 4 processors
</pre>
 
== GUI Access ==
 
 
===X11 forwarding===
Access to a GUI running on the cluster may be accomplished with X11 forwarding.  Data from the remote application is sent over ssh to an X server running locally.  Each additional ssh connection between the local machine and the cluster must be started with X11 forwarding enabled.  To request an interactive shell with X11 forwarding, you can use the "--x11" option. The following steps assume that the local machine has an X server running.
 
<ol>
<li> Login via ssh to front1 or front2.  Make sure your local ssh client has X11 forwarding enabled.  If you are using ssh on the command line, add the "-X" flag to your ssh command. </li>
<li> Start an interactive session with X11 forwarding.  Be sure to request the number of cores, amount of memory, and walltime to complete your job.  Syntax: <br/>
<pre>srun --x11 -N 1 -n 2 --mem=4096mb -t 8:00:00 --pty /bin/bash</pre></li>
<li>After starting an interactive session with X11 forwarding, you can now launch graphical programs from the terminal.</li>
</ol>
 
===RDP===
 
Remote Desktop Protocol provides a user with a graphical interface to connect to another computer over a network connection.  You will need a RDP client installed and will need to be connected to the KU Anywhere VPN.  You can RDP to either front1.ittc.ku.edu or front2.ittc.ku.edu.
 
== General Cluster Information ==
=== Software Environment ===
All cluster nodes run Rocky Linux 9.4 with GCC version 11.4.1. Cluster applications are installed as modules in /moosefs/apps/9/arch/generic/spack/modulefiles/linux-rocky9-x86_64.
 
==== Environment Modules ====
Cluster software is made available through environment modules. A list of available modules can be viewed by running:
 
<pre>
module avail
</pre>
 
Modules shown in the list can be loaded with the following command:
 
<pre>
module load module_name
</pre>
 
In order to persist loaded modules between interactive sessions, you need to add <code>module load</code> commands for the applications you want loaded to your ~/.bash_profile file if you are using bash, or ~/.cshrc if you are using tcsh or csh.
 
To view all loaded modules in your current shell session, use the <code>module list</code> command. To unload all currently loaded modules, you can use the <code>module purge</code> command. For more information on the module command and its options, see [https://linux.die.net/man/1/module the documentation] for further detail.
 
=== Filesystems ===
 
Below is a list of filesystems available on the cluster:
 
{| class="wikitable"
!|Path
!|Description
!|Default Quota
|-
|/home
|Personal Storage assigned to every user.
|50GB
|-
|/scratch
|Private working storage to run cluster jobs.
|1TB
|-
|/work
|Shared group storage.
|1TB
|-
|/users
|Stores private home directories. Avoid running cluster jobs out of this directory.
|5GB
|-
|/tmp
|Local storage on cluster nodes. (purged at end of job)
|N/A
|}
 
=== Debugging ===
 
The cluster has a number of tools at your disposal for debugging submitted Slurm jobs. The most basic debugging information available is from the log files generated by running your job, which contain the STDERR and STDOUT output from the job. Log files are located within the submit directory with the filename <code>slurm-<job id>.out</code>, such as <code>slurm-49321.out</code>.
 
You can retrieve detailed job information using the command <code>scontrol show jobid -dd <jobid></code>. Likewise, if you want to view detailed job information while the job is running, add the <code>--output</code> option to srun in your job batch file. For an unbuffered stream of STDOUT, which is quite useful for debugging, add the <code>-u</code> or <code>--unbuffered</code> to srun in your job batch file.
 
<!-- == Profiling == -->
 
=== Helpful Commands ===
The Slurm scheduler has a number of utilities for finding information on the status of your jobs. Below are listed a few of the most useful commands and options for quickly finding this information.
 
; Useful Slurm commands<b>:</b>:
:;sacct : Lists information on finished and currently running jobs, including job status and exit codes.
:;sacct -u <username> : Lists information on currently running and recently finished jobs for the specified user.
:;sacct -S <start-date> -s <state> : Lists all jobs that started before the start date or time that are in the specified state.
:;scancel -u <username> -t <state> : Cancels all of the jobs for the specific user that are in the specified state.
:;scontrol hold <jobid> : Suspends the specifed job by putting it in a 'HOLD' state.
:;scontrol resume <jobid> : Resumes the specified job from the 'HOLD' state.
:;scontrol show job <jobid> : Shows detailed queue and resource allocation information for the specified job.
:;sinfo : Displays information on all of the cluster partitions, including the nodes available in them.
:;sinfo -T : Shows information on cluster node reservations, including reservation period, name, and reserved nodes.
:;squeue : Displays the short-form information for all currently running and queued jobs.
:;squeue -u <username> -l : Lists the long-form information about currently running jobs for a specific user.
:;squeue -u <username> -t <state> : Lists information about a specific users jobs that are in the specified state.
:;sview : If X11 forwarding is enabled, this command launches a graphical interface for viewing cluster information.
 
== MMICC NVIDIA L40S GPU node ==
=== Requirement ===
You must have an ITTC/I2S Account and be part of the i2s-mmicc-gpu Group. Suzanne Michelle Shontz manages the group.
=== Submitting Jobs on MMICC NVIDIA L40S GPU node  ===
 
Instead of using hardware constraints, GPU L40 GPUs are specified with Generic Resource (gres) requests. Below is an example of an interactive GPU job request on mmicc partition with one L40s GPU:
<pre>
srun -p mmicc -c 8 --mem 16G --gres="gpu:l40s:1" --pty /bin/bash
</pre>
 
;The options used in this example are all detailed below<b>:</b>:
:; -p        : Specifies a partition
:; -c        : Sets the number of requested cpus per task
:; --mem    : This specifies the requested memory per node. Memory amounts can be given in Kilobytes (K), Megabytes (M), and Gigabytes (G).
:; --gres    : option allows the specification of a the GPU model and number through a colon-delimited list
:; --pty    : This option puts the srun session in pseudo-terminal mode. It is recommended to use this option if you are running an interactive shell session.
:; /bin/bash : The last option in an srun invocation is the program that srun will execute on the requested node. In this case, bash is specified to start an interactive shell session.
 
Note that that the default time limit for a job in the MMICC partition is 3 hours. Jobs that run past their allotted time will be stopped. To request more time, you'll need to use the -t parameter. This command will create a job in the mmicc partition with an 8 hour time limit:
<pre>
srun -c 8 --gres=gpu:l40s:1 --mem=8G -p mmicc -t 08:00:00 --pty /bin/bash
</pre>
 
You can verify your new time limit like this:
<pre>
[username@g030 ~]$ squeue --long
Thu Oct 17 15:42:06 2024
            JOBID PARTITION    NAME    USER    STATE      TIME TIME_LIMI  NODES NODELIST(REASON)
            46093    mmicc    bash username  RUNNING      1:34  8:00:00      1 g030
</pre>
 
 
The GPU partition must be specified when requesting GPUs, otherwise the scheduler will reject the job. Whenever a job is started on a GPU node, the environment variable CUDA_VISIBLE_DEVICES is set to contain a comma-delimited list of the GPUs allocated to the current job. Information about these GPUs can be viewed by running <code>nvidia-smi</code>.
Here is example output from the <code>srun</code> example above:
<pre>
[username@front1 scratch]$ srun -p mmicc -c 8 --mem 16G --gres="gpu:l40s:2" --pty /bin/bash
[username@g030 scratch]$ echo $CUDA_VISIBLE_DEVICES
0,1
[username@g030 scratch]$ nvidia-smi
Tue Oct 15 22:33:50 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 560.28.03              Driver Version: 560.28.03      CUDA Version: 12.6    |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf          Pwr:Usage/Cap |          Memory-Usage | GPU-Util  Compute M. |
|                                        |                        |              MIG M. |
|=========================================+========================+======================|
|  0  NVIDIA L40S                    On  |  00000000:0D:00.0 Off |                    0 |
| N/A  24C    P8            22W /  350W |      1MiB /  46068MiB |      0%      Default |
|                                        |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|  1  NVIDIA L40S                    On  |  00000000:B5:00.0 Off |                    0 |
| N/A  26C    P8            24W /  350W |      1MiB /  46068MiB |      0%      Default |
|                                        |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
 
+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU  GI  CI        PID  Type  Process name                              GPU Memory |
|        ID  ID                                                              Usage      |
|=========================================================================================|
|  No running processes found                                                            |
+-----------------------------------------------------------------------------------------+
</pre>
 
=== sbatch on MMICC NVIDIA L40S GPU node===
 
sbatch is used to submit jobs to the cluster using a script file. The --gres option allows the specification of the L40 GPU and number through a colon-delimited list. Below is a job script example:
<pre>
#!/bin/bash
#SBATCH -p mmicc
#SBATCH --gres=gpu:l40s:1
#SBATCH -c 4
#SBATCH --mem=8GB
#SBATCH -t 00:20:00
#SBATCH -J test_job
#SBATCH -o slurm-%j.out
echo "Job ${SLURM_JOB_ID} ran on ${HOSTNAME}"
</pre>
Example Output
<pre>
[username@front1 ~]$ sbatch test_job.sh
[username@front1 ~]$ cat slurm-47491.out
Job 47491 ran on g030
</pre>
 
== Cluster Hardware==
 
Visit the [[Cluster Hardware]] page for a complete listing of all of the nodes in the cluster and their hardware configurations.

Latest revision as of 10:17, 4 December 2024

Redirect to:

Retrieved from "https://help.ittc.ku.edu/index.php?title=Cluster_Documentation_2024&oldid=1808"