Difference between revisions of "Getting Started"

From UFRC
Jump to navigation Jump to search
 
(248 intermediate revisions by 20 users not shown)
Line 1: Line 1:
 +
[[Category:Essentials]]
 
{|align=right
 
{|align=right
 
   |__TOC__
 
   |__TOC__
 
   |}
 
   |}
==Getting an account==
+
Welcome to UF Research Computing! This page is intended to help new clients get started on HiPerGator.
To get an account at the UF HPC Center, you need to put a request in at our request page. That page is [http://www.hpc.ufl.edu/index.php?body=authuser located here].
 
==Logging In==
 
To login to the cluster, you need an SSH client of some sort. If you are using a linux or unix based system, there is most likely one already available to you in a shell, and you can get to your account very quickly.
 
  
If you are having problems connecting to your account, please let the HPC Staff know by submitting a [http://bugzilla.hpc.ufl.edu Bugzilla Request].
+
Note that new users are highly encouraged to take the [[New_user_training|New User Training]] course to familiarize themselves with using a supercomputer for research, save time, and avoid many beginner mistakes.
===Linux / Unix===
+
 
So, here is how you would go about logging in via a linux or unix account:
+
=From Zero to HiPerGator=
{|
+
==Initial Consult==
|
+
If a face-to-face discussion about the group's needs is needed you can [https://www.rc.ufl.edu/get-support/walk-in-support/ meet one of the UF Research Computing Facilitators] face-to-face or virtually or [https://support.rc.ufl.edu/ submit a support request to start the conversation].
<pre>
+
 
test@puppy:~$ ssh test@submit.hpc.ufl.edu
+
==HiPerGator Accounts==
test@submit.hpc.ufl.edu's password:
+
Group's sponsor has to be the first person to [https://www.rc.ufl.edu/access/account-request/ request a HiPerGator account] indicating that they are a new sponsor. In the process we will create their sponsored group.
Last login: Fri Feb  9 00:03:38 2007 from wsip-70-168-187-166.ga.at.cox.net
+
 
[test@submit ~]$
+
Afterwards, group members will be able to [https://www.rc.ufl.edu/access/account-request/ submit HiPerGator account requests] indicating their PI as the sponsor. Once approved, their linux accounts will be created.
</pre>
+
 
|}
+
==Trial Allocation==
The command <code>ssh test@submit.hpc.ufl.edu</code> is what you would type in at a command prompt on your system. After this, it asks you for a password, which you type in. After that, you are logged in and ready to work.
+
We recommend that the group's sponsor [https://gravity.rc.ufl.edu/access/request-trial-allocation/ request a '''free''' trial allocation] for storage and computational resources to get the group started on HiPerGator. Group members can then use HiPerGator for the 3 month duration of the trial allocation to figure out what resources and applications they really need.
 +
 
 +
==Purchasing Resources==
 +
After or while the group uses a trial allocation to determine the computational and storage resources it needs the group's sponsor can submit a purchase request for [https://gravity.rc.ufl.edu/access/purchase-request/hpg-hardware/ hardware (5-years)] or [https://gravity.rc.ufl.edu/access/purchase-request/hpg-service/ services (3-months to longer)] to invest into the resources to cover the group's HiPerGator use.
 +
 
 +
Some groups may have access to shared departmental allocations. In this case, instead of purchasing resources, group members can [https://support.rc.ufl.edu/ request] to be added to the departmental group to gain access to the shared resources.
 +
 
 +
Some examples of departments with shared allocations include the [http://ufgi.ufl.edu/ Genetics Institute], [http://epi.ufl.edu/ Emerging Pathogens Institute], [https://stat.ufl.edu/ Statistics Department], [http://biostat.ufl.edu/ Biostatistics Department], [https://www.eng.ufl.edu/ccmt/ Center for Compressible Multiphase Turbulence (CCMT)], [https://chp.phhp.ufl.edu/research/affiliated-centers/center-for-cognitive-aging-memory-cam/ Cognitive Aging and Memory Clinical Translational Research Program (CAMCTRP)], [https://efrc.ufl.edu/ Center for Molecular Magnetic Quantum Materials], [https://www.phys.ufl.edu/ Physics Department], and [https://plantpath.ifas.ufl.edu/ Plant Pathology Department]. In addition, several research groups working on collaborative projects have shared allocations accessible to members of those projects.
 +
 
 +
'''At this point a group is established on HiPerGator and can continue their computational work. See below for more details on the basic use.'''
 +
 
 +
=Introduction to Using HiPerGator=
 +
 
 +
To use HiPerGator or HiPerGator-AI you need three basic parts
 +
* Interfaces
 +
You use Interfaces to interact with the system, manage data, initialize computation, and view the results. The main categories of interfaces 'Command-Line' also known as Terminal, Graphical User Interfaces, and Web Interfaces or applications for more specialized use. Some distinctions here are blurred because, for example, you can open a Terminal while using a Web Interface like [[Jupyter#JupyterHub|JupyterHub]] or [[Open OnDemand]], but mostly you use a command-line Terminal interface through SSH connections (see below).
 +
* Data Management
 +
To perform research analyses you need to [[Transfer_Data|upload]] and [[Storage|manage]] data. Note that misuse of the storage systems is the second main reason for account suspension after running analyses on login nodes.
 +
*Computation
 +
'''Warning:''' do not run full-scale (normal) analyses on login nodes. [[Development and Testing]] is required reading. The main approach to run computational analyses is through writing [[Sample SLURM Scripts|job scripts]] and sending them to the [[SLURM_Commands|scheduler]] to run. Some interfaces like [[Open OnDemand]], [[Jupyter#JupyterHub|JupyterHub]], and [[Galaxy]] can manage job scheduling behind the scenes and may be more convenient than job submission from the command-line when appropriate.
 +
 
 +
For more information about on how to get started using HiPerGator visit our Wiki category  [https://help.rc.ufl.edu/doc/Category:Essentials Essentials], where you can find additional instructions, training, and tutorial videos.
 +
 
 +
==Interfaces==
 +
===Connecting to a HiPerGator Terminal via SSH===
 +
To work on HiPerGator you will have to connect to hpg.rc.ufl.edu from your local computer either via SSH (terminal session) or via one of the web/application interfaces we provide such as [[Galaxy]], [[Open_OnDemand|Open OnDemand]], or [[Jupyter#JupyterHub|JupyterHub]]. The default port 22 will require password authentication whereas port 2222 can be used for ssh key authentication.
 +
 
 +
Note that the UF security requirements include the use of [[Authentication_With_MFA|Multi-Factor Authentication]] when connecting to HiPerGator. See the [[SSH Multiplexing]] help if you are connecting from Linux or MacOS to reduce the frequency of MFA use.
 +
 
 +
For any given command below, <code><username></code> should be replaced with the UFRC username (same as your GatorLink username).
 +
 
 +
====Connecting from Windows====
 +
<div class="mw-collapsible mw-collapsed" style="width:70%; padding: 5px; border: 1px solid gray;">
 +
''Expand this section to view instructions for logging in with Windows.''
 +
<div class="mw-collapsible-content" style="padding: 5px;">
 +
Since Microsoft Windows does not come with a built-in SSH client, you must download a client from the web.
 +
 
 +
For University-managed computers [http://www.chiark.greenend.org.uk/~sgtatham/putty PuTTY], [https://tabby.sh/ Tabby], and [https://gitforwindows.org/ Git Bash] are approved for 'fast track' installations.
 +
 
 +
'''PuTTY'''
 +
* [http://www.chiark.greenend.org.uk/~sgtatham/putty Download PuTTY] to your local machine and start the program.
 +
* Connect to hpg.rc.ufl.edu.
 +
* At the login prompt, enter your username (this should be the same as your GatorLink username)
 +
* Enter your password when prompted. You are now connected and ready to work!
 +
 
 +
'''Tabby'''
 +
* [https://github.com/Eugeny/tabby/releases/latest Download Tabby] to your local machine: tabby-version#-setup.exe or tabby-version#-portable.zip for a portable version.
 +
* Start the program and click Settings > Profiles > +New profile > SSH connection
 +
  Name: HiPerGator
 +
  Host: hpg.rc.ufl.edu
 +
  Username: <username>
 +
  Password: "Set password" or "Add a private key"
 +
* Click "Save"
 +
* Click on the window icon "New tab with profile" and select "HiPerGator hpg.rc.ufl.edu"
 +
* You are now connected and ready to work!
 +
</div>
 +
</div>
 +
 
 +
====Connecting from Linux and MacOS====
 +
<div class="mw-collapsible mw-collapsed" style="width:70%; padding: 5px; border: 1px solid gray;">
 +
''Expand to view instructions for connecting from Linux or MacOS.''
 +
<div class="mw-collapsible-content" style="padding: 5px;">
 +
You may omit GatorLinkUsername if the username on your local computer matches your GatorLink username e.g.
 +
 
 +
Open a terminal and run
 +
$ ssh <GatorLink Username>@hpg.rc.ufl.edu
 +
or
 +
$ ssh hpg.rc.ufl.edu
 +
 
 +
Enter your password when the prompt appears. You are now connected to a login node and your working directory will be <code>/home/<GatorLink></code>, ready to work!
 +
</div>
 +
</div>
 +
 
 +
===Logging in via web browser===
 +
<div class="mw-collapsible mw-collapsed" style="width:70%; padding: 5px; border: 1px solid gray;">
 +
''Expand to view instructions for Logging in via web browser.''
 +
<div class="mw-collapsible-content" style="padding: 5px;">
 +
* To start a GUI application go to [https://ood.rc.ufl.edu https://ood.rc.ufl.edu] and choose the application you would like to use, for example HiPerGator Desktop is a linux GUI.
 +
**
 +
** To start a Jupyter Notebook with OOD, select the “Jupyter Notebook” application server under the “Interactive Applications” tab on the main dashboard. Specify the resources you need in the resource request and click “Launch” at the bottom of the form. Once your job starts, click “Connect to Jupyter” to start your Notebook.
 +
*
 +
* To start a Jupyter Notebook go to [http://jhub.rc.ufl.edu http://jhub.rc.ufl.edu] and select the resources you need. This JupyterHub server has preset resource options.
 +
**
 +
** Jupyterhub and OutOfMemory (OOM) – if your Notebook stops responding it might be an indication of the job running out of memory. Currently this does not produce an error message. You can check your <code>/home/<user>/</code>jupyter slurm spawner logs to be sure. You will need to restart the job with additional RAM.
 +
*
 +
* To start a Galaxy instance go to [https://galaxy.rc.ufl.edu/ https://galaxy.rc.ufl.edu/]. Galaxy is a web-based framework for accessible, reproducible, and transparent biological computing.
 +
</div>
 +
</div>
 +
 
 +
===Invalid Password===
 +
If you have forgotten your password or need to change it, use the following link to set a new one: [https://my.ufl.edu/ps/signon.html Change or Reset Password]
 +
 
 +
'''NOTE''': Research Computing now uses Gatorlink Authentication for all passwords. Please use the UF Gatorlink password system for all password issues.
 +
 
 +
==Data Management==
 +
 
 +
===Transferring Data===
 +
If you need to transfer datasets to or from HiPerGator and your local computer or another external location you have to pick the appropriate transfer mechanism.
 +
* For small or medium file transfers use sftp, scp, or rsync to login to sftp.rc.ufl.edu, hpg.rc.ufl.edu, or rsync.rc.ufl.edu.
 +
*
 +
* For large file transfers or transfers with many small files, use our [[Globus]] service.
 +
 
 +
For more in-depth information see [[Transfer_Data|Transfer Data]].
 +
 
 +
====SFTP====
 +
SFTP, or secure file transfer, works well for small to medium data transfers and is appropriate for both small and large data files.
 +
 
 +
If you would like to use a Graphical User Interface secure file transfer client we recommend:
 +
* [https://tabby.sh/ Tabby] for Windows, Linux, or MacOS.
 +
* [https://mobaxterm.mobatek.net/ MobaXterm] for Windows.
 +
* [http://winscp.net/eng/index.php WinSCP] for Windows.
 +
* [http://cyberduck.io/ Cyberduck] for MacOS X and Windows.
 +
 
 +
After you have chosen and downloaded a client, configure the client to connect to <code>hpg.rc.ufl.edu</code>, specifying port number 22. Use your username and password to log in.
 +
 
 +
====Rsync====
 +
If you prefer to use the command-line or to get maximum efficiency from your data transfers Rsync, which is an incremental file transfer utility that minimizes network usage, is a good choice. It does so by transmitting only the differences between local and remote files rather than transmitting complete files every time a sync is run as SFTP does. Rsync is best used for tasks like synchronizing files stored across multiple subdirectories, or updating large data sets. It works well both for small and large files. [[Rsync|See the Rsync page]] for instructions on using rsync.
 +
 
 +
===Globus===
 +
Globus is a high-performance mechanism for file transfer. Globus works especially well for transferring large files or data sets
 +
* [[Globus|See the Globus page]] for setup and configuration information.
 +
 
 +
===Samba===
 +
Samba service, also known as a '<code>network share</code>' or '<code>mapped drive</code>' provides you with an ability to connect to some HiPerGator filesystems as locally mapped drives (or mount points on Linux or MacOS X).Once you connected to a share this mechanism provides you with a file transfer option that allows you to use your client computer's native file manager to access and manage your files.  UFRC Samba setup does not provide high performance, so try to use it sparingly and for smaller files, like job scripts or analysis reports to be copied to  your local system. You must be connected to the UF network (either on-campus or through the UF VPN) to connect to Samba shares.
  
===Windows===
+
* [[Samba_Access|See the page on accessing Samba]] for setup information specific to your computer's operating system.
For Microsoft Windows, things are a bit trickier. Windows does not come with a built-in SSH client, and this makes things difficult. What you have to do is download a client from the Internet and install it, then use that client. We recommend the following:
 
* For a shell client, [http://www.chiark.greenend.org.uk/~sgtatham/putty Putty]
 
* For a file transfer client, [http://winscp.net/eng/index.php WinSCP]
 
Both of the above clients have documentation at their websites, so I will not go into it here. Once you are logged in and have a prompt that resembles that in the Unix / Linux section above, you can continue with this tutorial.
 
===Passwords===
 
The passwords assigned to users at the HPC Center expire every six months. If you receive a message stating that it is about to expire, you cna reset it by using the following commands:
 
<pre>
 
$ passwd
 
Enter new UNIX password:
 
Retype new UNIX password:
 
</pre>
 
As you can see above, the system will ask for the password twice as it wants to confirm that you have typed it in correctly. Nothing will be displayed when you are typing the actual password.
 
  
==Changing your Password==
+
===Automounted Paths===
In order to change your password to something perhaps a little easier to remember than the password supplied, which typically looks like a line of static, you can do the following:
+
Note: NFS-based storage on our systems are typically automounted, which means they are dynamically mounted only when users are actually accessing them. For example if you have an invested folder as /orange/smith, to access it you will have to specifically type in the full path of "/orange/smith" to be able to see the contents and access them. Directly browsing /orange will not show the smith sub-folder unless someone else is using it coincidentally. Automounted folders are pretty common on the systems, they include /orange, /bio, /rlts and even /home etc.
<pre>
 
$ passwd
 
Changing password for user <YOUR USERNAME>.
 
Enter login(LDAP) password:
 
New UNIX password:
 
Retype new UNIX password:
 
New password:
 
Re-enter new password:
 
LDAP password information changed for <YOUR USERNAME>
 
passwd: all authentication tokens updated successfully.
 
</pre>
 
The above asks for your old password, then asks for your new password a total of four times. This seems like a lot, but it is actually a nice thing because it does the following:
 
* Makes sure that you are entering your new password correctly.
 
* Does a bit of training your fingers for muscle memory.
 
==Looking Around==
 
We expect the users of the HPC Center cluster to already have a working knowledge of the linux operating system, so we are not going to go into detail here on using the operating system. Below are some links to webpages that describe a lot of this, however:
 
* [http://www.tutorialized.com/tutorial/Basic-Linux-Shell-commands/21596 Basic Linux Shell Commands]
 
  
===File System===
+
==Editing your files==
We have a structured file system that is important to know about. Please read about it here: [[HPC File System]]
+
Several methods exist for editing your files on the cluster.
 +
===Native Editors===
 +
* '''vi''' - The visual editor (vi) is the traditional Unix editor; however, it is not necessarily the most intuitive editor. [http://www.eng.hawaii.edu/Tutor/vi.html View a tutorial for using vi]
 +
* '''emacs''' - Emacs is a much heavier duty editor, but again has the problem of having commands that are non-intuitive. [http://www2.lib.uchicago.edu/~keith//tcl-course/emacs-tutorial.html View a tutorial for using emacs]
 +
* '''nano''' - Nano has a good bit of on-screen help to make it easier to use.
  
==Editing==
 
Editing files on the cluster can be done through a couple of different methods...
 
===In-System Editors===
 
* vi - vi is the basic editor for a number of people. Using the editor is not necessarily intuitive, so we have provided a link to a tutorial.
 
** [http://www.eng.hawaii.edu/Tutor/vi.html VI Tutorial]
 
* emacs - emacs is a much heavier duty editor, but again has the problem of having commands that are non-intuitive. Again, we have provided a link to a tutorial for this editor.
 
** [http://www2.lib.uchicago.edu/~keith//tcl-course/emacs-tutorial.html Emacs Tutorial]
 
 
===External Editors===
 
===External Editors===
You can also use your favorite editor on your local machine and then transfer the files over to the HPC Cluster afterwards. One caveat to this is that with files created on Windows machines, very often extra characters get injected into text files which can cause major problems when it comes time to interpret the files. If this happens, there is a utility called <code>dos2unix</code> that you can use to remove the extra characters.
+
You can also use your favorite file editor on your local machine, and then transfer the files to the cluster afterward. A caveat to this is that files created on Windows machines usually contain unprintable characters, which may be misinterpreted by Linux command interpreters (shells). If this happens, there is a utility called <code>dos2unix</code> that you can use to convert the text file from DOS/Windows formatting to Linux formatting.
 +
 
 +
==Computation==
 +
===HiPerGator Etiquette===
 +
* '''Only run workloads on compute nodes.''' Do not run scripts or applications on the login nodes beyond a [[Development_and_Testing#Login_Nodes|small quick test]]. Use <code>sbatch</code>, <code>srundev</code>, <code>salloc</code>, or <code>srun</code> to start a session on a compute node instead.
  
==Running Jobs==
+
* '''Only run workloads from blue storage.''' This is a fast storage systems that can handle the I/O involved in research workloads. Before using <code>sbatch</code> or launching a workload interactively, make sure your working directory is a blue file path, e.g. <code>/blue/<group>/<user></code>, and not your home directory (<code>~</code> or <code>/home/<user></code>). Use <code>pwd</code> to print working directory.  
===General Scheduling===
 
Jobs from faculty investors in the HPC Center are now favored over jobs from groups who did not invest in the HPC Center.
 
  
Job scheduling has been a big topic with the HPC committee in the last several months. The HPC Center staff has been directed by the committee to improve the quality of service of job scheduling for jobs coming from investors in the HPC Center.  This means reducing the time spent in the queues and allowing jobs from the investors to capture the full share of the resources that they have paid for.  The HPC committee recently adopted a document which spells out what they want.
+
* '''Home directory is only for storing user readable files.''' Your 40GB home directory is the only storage on HiPerGator for which file recovery may be possible, so keep copies of scripts, configurations,  or other important files here. Versions of files '''may''' be available for the previous 7 days ([[Snapshots|see information on home directory snapshots and recovering files]]), if you need to recover a file from your home directory. If you need back ups of important data, backup services will need to be purchased.  
  
Jobs can be submitted on '''submit.hpc.ufl.edu'''.
+
* '''Do not install new software when using existing modules.''' This will cause errors when using our software environments, because the new installation is in your local folder. If you need to install software, create a [[Conda]] virtual environment, or open a support ticket to request RC to install the software in one of our software environments if it will be widely used.
  
===Torque Scheduler===
+
* '''Do not run workloads from orange storage.''' Orange is intended as long-term, archival storage of data you currently do not use. It cannot handle the high-throughput requirements of high-performance computing workloads.  
The [http://www.clusterresources.com/pages/products/torque-resource-manager.php Torque Resource Manager] has been installed on the HPC cluster, and is slowly being switched over to from the PBS Pro scheduler as time goes on. The [[Maui]] scheduler is also running in this environment.
 
  
The Torque scheduler is installed on the cluster, and accepts the same commands as the PBS Pro scheduler with a couple of exceptions. Currently we are experimenting with these packages so that we can provide improved scheduling to HPC users. While we are still learning about Torque and [[Maui]], our experiences so far have been good and we are guardedly optimistic that Torque and Maui will end up being the resource manager and scheduler for the HPC Center sometime in the not-too-distant future.
+
* '''Do not request excessive resources.''' This includes CPU, GPU and memory. Job emails include summary estimations of memory use, however, active monitoring will help you understand resource requirements. Applications often require special commands, arguments, or configurations to run in parallel. Therefore, you will likely need to do more than request multiple CPUs or GPUs for a workload to put those resources to use.
  
Please note the following.
+
===Using installed software===
 +
The full list of software available for use can be viewed on the [[Installed_Software|Installed Software]] page. Access to installed software is provided through [[Modules|Environment Modules]].
  
* If your job is single-threaded (1 cpu) and does not have heavy I/O requirements, it does not need to run on an infiniband-enabled node.  In that case, you should include the "gige" property in your PBS resource specification as follows
+
The following command can be used to browse the full list of available modules, along with short descriptions of the applications they make available:
 
<pre>
 
<pre>
#PBS  -l nodes=1:ppn=1:gige
+
module spider
 
</pre>
 
</pre>
* If you need to run an mpi-based application that has not been rebuilt for OpenMPI 1.2.0+Torque, please send us a note and we'll be happy to rebuild  what you need - first come, first serve.
 
* If you build your own MPI-based application executables, you should use the MPI compiler wrappers (mpif90, mpicc, mpiCC) in /opt/intel/ompi/1.2.0/bin.  These wrappers will automatically pull in the correct libraries.
 
* We will continue to tune the maui scheduler to provide fair and efficient scheduling according to the policies established by the HPC Committee and within the capabilities of the maui scheduler.    Keep in mind that these policies include priority and quality-of-service commitments to those faculty who have invested in the resources within the HPC Center.
 
  
 +
To load a module, use the following command:
 +
<pre>
 +
module load <module_name>
 +
</pre>
 +
 +
In Jupyter Notebooks, kernels are available with our most popular software stacks. If you are unable to find what you need or would like software installed, please fill out a [https://www.rc.ufl.edu/help/support-requests/ help request].
 +
 +
 +
For more information on loading modules to access software, view the page on the [[Modules_Basic_Usage|basic usage of environment modules]].
 +
 +
There are some useful commands and utilities in a [[UFRC_environment_module|'ufrc' environment module]] in addition to installed applications.
 +
 +
===Interactive Testing or Development===
 +
You don't always have to use the SLURM scheduler. When all you need is a quick shell session to run a command or two, write and/or test a job script, or compile some code use [[Development_and_Testing|SLURM Dev Sessions]].
 +
 +
===Running Graphical Programs===
 +
It is possible to run programs that use a graphical user interface (GUI) on the system. However, doing so requires an installation of and configuration of additional software on the client computer.
 +
 +
Please see the [[GUI_Programs|GUI Programs]] page for information on running graphical user interface applications at UFRC.
 +
 +
===Scheduling computational jobs===
 +
UFRC uses the Simple Linux Utility for Resource Management, or '''SLURM''', to allocate resources and schedule jobs. Users can create SLURM job scripts to submit jobs to the system. These scripts can, and should, be modified in order to control several aspects of your job, like resource allocation, email notifications, or an output destination.
  
 +
* See the [[Annotated SLURM Script]] for a walk-through of the basic components of a SLURM job script
 +
* See the [[Sample SLURM Scripts]] for several SLURM job script examples
  
====Trivial Example====
 
{|
 
|-valign="top"
 
|
 
<pre>
 
#! /bin/sh
 
#PBS -N testjob
 
#PBS -o testjob.out
 
#PBS -e testjob.err
 
#PBS -M mxcheng@ufl.edu
 
#PBS -l walltime=00:01:00
 
#PBS -l nodes=1:ppn=1
 
  
date
+
To submit a job script from one of the login nodes accessed via hpg.rc.ufl.edu, use the following command:
hostname
 
</pre>
 
||
 
To submit this job from [[submit.hpc.ufl.edu]], you would use the following command:
 
 
<pre>
 
<pre>
$ qsub <your job script>
+
$ sbatch <your_job_script>
 
</pre>
 
</pre>
To check the status of running jobs, you would use the following command:
+
To check the status of submitted jobs, use the following command:
 
<pre>
 
<pre>
$ qstat [-u <username>]
+
$ squeue -u <username>
 
</pre>
 
</pre>
or [http://www.hpc.ufl.edu/ HPC] --> Utilization --> [http://www.hpc.ufl.edu/index2.php?body=queue2 Torque Queue Status]
 
|}
 
  
==Compiling your own==
+
View [[SLURM Commands]] for more useful SLURM commands.
By default, when you first login to the system you have compilers setup for the Intel OpenMPI compilers. This gives you access to C, C++, F77 and F90 compilers in your path, as well as the mpicc and mpif90 compilers for OpenMPI applications. If you want to change this, you can use a selector program on the cluster called "mpi-selector". The command specification for this is as follows:
+
 
<pre>
+
====Managing Cores and Memory====
/usr/bin/mpi-selector options:
+
See [[Account and QOS limits under SLURM]] for the main documentation on efficient management of computational resources, and an extensive explanation of QOS and SLURM account use.
 +
 
 +
The amount of resources within an investment is calculated in NCU (Normalized Computing Units), which correspond to 1 CPU core and about 3.5GB of memory for each NCU purchased. CPUs (cores) and RAM are allocated to jobs independently as requested by your job script.
 +
 
 +
Your group's investment can run out of **cores** (SLURM may show <code>QOSGrpCpuLimit</code> in the reason a job is pending) OR **memory** (SLURM may show <code>QOSGrpMemLimit</code> in the reason a job is pending) depending on current use by running jobs.
 +
The majority of HiPerGator nodes have the same ratio of about 4 GB of RAM per core, which, after accounting for the operating system and system services, leaves about 3.5 GB usable for jobs; hence the ratio of 1 core and 3.5GB RAM per NCU.
 +
 
 +
Most HiPerGator nodes have 32 cores and 128 GB RAM (~30,000 cores in the newer part of the cluster) or 64 cores and 256 GB RAM (~16,000 cores in the older part of the cluster). The '''bigmem''' nodes and the newer Skylake nodes have a higher ratio of 16 GB/core and 6 GB/core, respectively. See [[Available_Node_Features]] for the exact data on resources available on all types of nodes on HiPerGator.
  
Options for MPI implementations:
+
You must specify both the number of cores and the amount of RAM needed in the job script for SLURM with the <code>--mem</code> (total job memory) or <code>--mem-per-cpu</code> (per-core memory) options. Otherwise, the job will be assigned the default 600mb of memory.
 +
 +
If you need more than 128 GB of RAM, you can only run on the older nodes, which have 256 GB of RAM, or on the bigmem nodes, which have up to 1.5 TB of RAM.
  
--register <name>    Register an MPI implementation with the central
+
===Monitoring Your Workloads===
                      mpi-selector database. Requires use of the
+
You can see presently running workloads with the squeue command e.g.
                      --source-dir option.
+
  $ squeuemine
--source-dir <dir>    Used with --register, indicating that <dir> is
+
OpenOnDemand offers a method to monitor jobs using the Jobs menu in the upper toolbar on your dashboard. This will show your current running, pending, and recently completed jobs. Select: Jobs -> Active Jobs from the upper dashboard menu.
                      where mpivars.sh and mpivars.csh can be found.
 
--unregister <name>   Remove an MPI implementation list from the
 
                      central mpi-selector database.
 
  
Options for system adminsitrators:
 
  
--system              When used with the --set and --unset options,
+
We provide a number of helpful commands in the [[UFRC_environment_module|UFRC module]]. The <code>ufrc</code> module is loaded by default at login, but you can also load the <code>ufrc</code> module with the following command:
                      act on the system-wide defaults (vs. the
+
  $ module load ufrc
                      per-user defaults).  When used with --query, only
 
                      show the site-wide default (if there is one).
 
--user                When used with the --set and --unset options,
 
                      act on the per-user defaults (vs. the
 
                      site-wide defaults). When used with --query, only
 
                      show the per-user default (if there is one).
 
  
Options for users:
 
  
--list                Show a list of the currently registered MPI
+
Examples of commands for SLURM or HiPerGator specific UFRC environment module
                      implementations.
+
<div class="mw-collapsible mw-collapsed" style="width:70%; padding: 5px; border: 1px solid gray;">
--set <name>         Set <name> to be the default MPI selection.
+
''Expand this section to view Examples of commands''
--unset              Remove the default MPI selection.
+
<div class="mw-collapsible-content" style="padding: 5px;">
--query              Shows the current default MPI selection.
+
<pre>
--yes                Assume the answer is "yes" to any question.
+
$ slurmInfo          - displays resource usage for your group
--no                  Assume the answer is "no" to any question.
+
$ slurmInfo -p        - displays resource usage per partition
--verbose            Be verbose about actions.
+
$ showQos            - displays your available QoS
--silent              Print nothing (not even warnings or errors;
+
$ home_quota          - displays your /home quota
                      overrides --verbose)
+
$ blue_quota          - displays your /blue quota
--version            Display the version of /usr/bin/mpi-selector.
+
$ orange_quota        - displays your /orange quota
 +
$ sacct              - displays job id and state of your recent workloads
 +
$ nodeInfo            - displays partitions by node types, showing total RAM and other features
 +
$ sinfo -p partition  - displays the status of nodes in a partition
 +
$ jobhtop            - displays resource usage for jobs
 +
$ jobnvtop            - displays resource usage for GPU jobs
 +
$ which python        - displays path to the Python install of the environment modules you have loaded
 
</pre>
 
</pre>
 +
 +
</div>
 +
</div>
 +
 +
==Getting help==
 +
If you are having problems using the UFRC system, try our resources at [https://help.rc.ufl.edu/doc/Category:Help Help] or let our staff know by submitting a [http://support.rc.ufl.edu support request].

Latest revision as of 16:48, 23 November 2022

Welcome to UF Research Computing! This page is intended to help new clients get started on HiPerGator.

Note that new users are highly encouraged to take the New User Training course to familiarize themselves with using a supercomputer for research, save time, and avoid many beginner mistakes.

From Zero to HiPerGator

Initial Consult

If a face-to-face discussion about the group's needs is needed you can meet one of the UF Research Computing Facilitators face-to-face or virtually or submit a support request to start the conversation.

HiPerGator Accounts

Group's sponsor has to be the first person to request a HiPerGator account indicating that they are a new sponsor. In the process we will create their sponsored group.

Afterwards, group members will be able to submit HiPerGator account requests indicating their PI as the sponsor. Once approved, their linux accounts will be created.

Trial Allocation

We recommend that the group's sponsor request a free trial allocation for storage and computational resources to get the group started on HiPerGator. Group members can then use HiPerGator for the 3 month duration of the trial allocation to figure out what resources and applications they really need.

Purchasing Resources

After or while the group uses a trial allocation to determine the computational and storage resources it needs the group's sponsor can submit a purchase request for hardware (5-years) or services (3-months to longer) to invest into the resources to cover the group's HiPerGator use.

Some groups may have access to shared departmental allocations. In this case, instead of purchasing resources, group members can request to be added to the departmental group to gain access to the shared resources.

Some examples of departments with shared allocations include the Genetics Institute, Emerging Pathogens Institute, Statistics Department, Biostatistics Department, Center for Compressible Multiphase Turbulence (CCMT), Cognitive Aging and Memory Clinical Translational Research Program (CAMCTRP), Center for Molecular Magnetic Quantum Materials, Physics Department, and Plant Pathology Department. In addition, several research groups working on collaborative projects have shared allocations accessible to members of those projects.

At this point a group is established on HiPerGator and can continue their computational work. See below for more details on the basic use.

Introduction to Using HiPerGator

To use HiPerGator or HiPerGator-AI you need three basic parts

  • Interfaces

You use Interfaces to interact with the system, manage data, initialize computation, and view the results. The main categories of interfaces 'Command-Line' also known as Terminal, Graphical User Interfaces, and Web Interfaces or applications for more specialized use. Some distinctions here are blurred because, for example, you can open a Terminal while using a Web Interface like JupyterHub or Open OnDemand, but mostly you use a command-line Terminal interface through SSH connections (see below).

  • Data Management

To perform research analyses you need to upload and manage data. Note that misuse of the storage systems is the second main reason for account suspension after running analyses on login nodes.

  • Computation

Warning: do not run full-scale (normal) analyses on login nodes. Development and Testing is required reading. The main approach to run computational analyses is through writing job scripts and sending them to the scheduler to run. Some interfaces like Open OnDemand, JupyterHub, and Galaxy can manage job scheduling behind the scenes and may be more convenient than job submission from the command-line when appropriate.

For more information about on how to get started using HiPerGator visit our Wiki category Essentials, where you can find additional instructions, training, and tutorial videos.

Interfaces

Connecting to a HiPerGator Terminal via SSH

To work on HiPerGator you will have to connect to hpg.rc.ufl.edu from your local computer either via SSH (terminal session) or via one of the web/application interfaces we provide such as Galaxy, Open OnDemand, or JupyterHub. The default port 22 will require password authentication whereas port 2222 can be used for ssh key authentication.

Note that the UF security requirements include the use of Multi-Factor Authentication when connecting to HiPerGator. See the SSH Multiplexing help if you are connecting from Linux or MacOS to reduce the frequency of MFA use.

For any given command below, <username> should be replaced with the UFRC username (same as your GatorLink username).

Connecting from Windows

Expand this section to view instructions for logging in with Windows.

Since Microsoft Windows does not come with a built-in SSH client, you must download a client from the web.

For University-managed computers PuTTY, Tabby, and Git Bash are approved for 'fast track' installations.

PuTTY

  • Download PuTTY to your local machine and start the program.
  • Connect to hpg.rc.ufl.edu.
  • At the login prompt, enter your username (this should be the same as your GatorLink username)
  • Enter your password when prompted. You are now connected and ready to work!

Tabby

  • Download Tabby to your local machine: tabby-version#-setup.exe or tabby-version#-portable.zip for a portable version.
  • Start the program and click Settings > Profiles > +New profile > SSH connection
 Name: HiPerGator
 Host: hpg.rc.ufl.edu
 Username: <username>
 Password: "Set password" or "Add a private key"
  • Click "Save"
  • Click on the window icon "New tab with profile" and select "HiPerGator hpg.rc.ufl.edu"
  • You are now connected and ready to work!

Connecting from Linux and MacOS

Expand to view instructions for connecting from Linux or MacOS.

You may omit GatorLinkUsername if the username on your local computer matches your GatorLink username e.g.

Open a terminal and run

$ ssh <GatorLink Username>@hpg.rc.ufl.edu
or
$ ssh hpg.rc.ufl.edu

Enter your password when the prompt appears. You are now connected to a login node and your working directory will be /home/<GatorLink>, ready to work!

Logging in via web browser

Expand to view instructions for Logging in via web browser.

  • To start a GUI application go to https://ood.rc.ufl.edu and choose the application you would like to use, for example HiPerGator Desktop is a linux GUI.
    • To start a Jupyter Notebook with OOD, select the “Jupyter Notebook” application server under the “Interactive Applications” tab on the main dashboard. Specify the resources you need in the resource request and click “Launch” at the bottom of the form. Once your job starts, click “Connect to Jupyter” to start your Notebook.
  • To start a Jupyter Notebook go to http://jhub.rc.ufl.edu and select the resources you need. This JupyterHub server has preset resource options.
    • Jupyterhub and OutOfMemory (OOM) – if your Notebook stops responding it might be an indication of the job running out of memory. Currently this does not produce an error message. You can check your /home/<user>/jupyter slurm spawner logs to be sure. You will need to restart the job with additional RAM.
  • To start a Galaxy instance go to https://galaxy.rc.ufl.edu/. Galaxy is a web-based framework for accessible, reproducible, and transparent biological computing.

Invalid Password

If you have forgotten your password or need to change it, use the following link to set a new one: Change or Reset Password

NOTE: Research Computing now uses Gatorlink Authentication for all passwords. Please use the UF Gatorlink password system for all password issues.

Data Management

Transferring Data

If you need to transfer datasets to or from HiPerGator and your local computer or another external location you have to pick the appropriate transfer mechanism.

  • For small or medium file transfers use sftp, scp, or rsync to login to sftp.rc.ufl.edu, hpg.rc.ufl.edu, or rsync.rc.ufl.edu.
  • For large file transfers or transfers with many small files, use our Globus service.

For more in-depth information see Transfer Data.

SFTP

SFTP, or secure file transfer, works well for small to medium data transfers and is appropriate for both small and large data files.

If you would like to use a Graphical User Interface secure file transfer client we recommend:

After you have chosen and downloaded a client, configure the client to connect to hpg.rc.ufl.edu, specifying port number 22. Use your username and password to log in.

Rsync

If you prefer to use the command-line or to get maximum efficiency from your data transfers Rsync, which is an incremental file transfer utility that minimizes network usage, is a good choice. It does so by transmitting only the differences between local and remote files rather than transmitting complete files every time a sync is run as SFTP does. Rsync is best used for tasks like synchronizing files stored across multiple subdirectories, or updating large data sets. It works well both for small and large files. See the Rsync page for instructions on using rsync.

Globus

Globus is a high-performance mechanism for file transfer. Globus works especially well for transferring large files or data sets

Samba

Samba service, also known as a 'network share' or 'mapped drive' provides you with an ability to connect to some HiPerGator filesystems as locally mapped drives (or mount points on Linux or MacOS X).Once you connected to a share this mechanism provides you with a file transfer option that allows you to use your client computer's native file manager to access and manage your files. UFRC Samba setup does not provide high performance, so try to use it sparingly and for smaller files, like job scripts or analysis reports to be copied to your local system. You must be connected to the UF network (either on-campus or through the UF VPN) to connect to Samba shares.

Automounted Paths

Note: NFS-based storage on our systems are typically automounted, which means they are dynamically mounted only when users are actually accessing them. For example if you have an invested folder as /orange/smith, to access it you will have to specifically type in the full path of "/orange/smith" to be able to see the contents and access them. Directly browsing /orange will not show the smith sub-folder unless someone else is using it coincidentally. Automounted folders are pretty common on the systems, they include /orange, /bio, /rlts and even /home etc.

Editing your files

Several methods exist for editing your files on the cluster.

Native Editors

  • vi - The visual editor (vi) is the traditional Unix editor; however, it is not necessarily the most intuitive editor. View a tutorial for using vi
  • emacs - Emacs is a much heavier duty editor, but again has the problem of having commands that are non-intuitive. View a tutorial for using emacs
  • nano - Nano has a good bit of on-screen help to make it easier to use.

External Editors

You can also use your favorite file editor on your local machine, and then transfer the files to the cluster afterward. A caveat to this is that files created on Windows machines usually contain unprintable characters, which may be misinterpreted by Linux command interpreters (shells). If this happens, there is a utility called dos2unix that you can use to convert the text file from DOS/Windows formatting to Linux formatting.

Computation

HiPerGator Etiquette

  • Only run workloads on compute nodes. Do not run scripts or applications on the login nodes beyond a small quick test. Use sbatch, srundev, salloc, or srun to start a session on a compute node instead.
  • Only run workloads from blue storage. This is a fast storage systems that can handle the I/O involved in research workloads. Before using sbatch or launching a workload interactively, make sure your working directory is a blue file path, e.g. /blue/<group>/<user>, and not your home directory (~ or /home/<user>). Use pwd to print working directory.
  • Home directory is only for storing user readable files. Your 40GB home directory is the only storage on HiPerGator for which file recovery may be possible, so keep copies of scripts, configurations, or other important files here. Versions of files may be available for the previous 7 days (see information on home directory snapshots and recovering files), if you need to recover a file from your home directory. If you need back ups of important data, backup services will need to be purchased.
  • Do not install new software when using existing modules. This will cause errors when using our software environments, because the new installation is in your local folder. If you need to install software, create a Conda virtual environment, or open a support ticket to request RC to install the software in one of our software environments if it will be widely used.
  • Do not run workloads from orange storage. Orange is intended as long-term, archival storage of data you currently do not use. It cannot handle the high-throughput requirements of high-performance computing workloads.
  • Do not request excessive resources. This includes CPU, GPU and memory. Job emails include summary estimations of memory use, however, active monitoring will help you understand resource requirements. Applications often require special commands, arguments, or configurations to run in parallel. Therefore, you will likely need to do more than request multiple CPUs or GPUs for a workload to put those resources to use.

Using installed software

The full list of software available for use can be viewed on the Installed Software page. Access to installed software is provided through Environment Modules.

The following command can be used to browse the full list of available modules, along with short descriptions of the applications they make available:

module spider

To load a module, use the following command:

module load <module_name>

In Jupyter Notebooks, kernels are available with our most popular software stacks. If you are unable to find what you need or would like software installed, please fill out a help request.


For more information on loading modules to access software, view the page on the basic usage of environment modules.

There are some useful commands and utilities in a 'ufrc' environment module in addition to installed applications.

Interactive Testing or Development

You don't always have to use the SLURM scheduler. When all you need is a quick shell session to run a command or two, write and/or test a job script, or compile some code use SLURM Dev Sessions.

Running Graphical Programs

It is possible to run programs that use a graphical user interface (GUI) on the system. However, doing so requires an installation of and configuration of additional software on the client computer.

Please see the GUI Programs page for information on running graphical user interface applications at UFRC.

Scheduling computational jobs

UFRC uses the Simple Linux Utility for Resource Management, or SLURM, to allocate resources and schedule jobs. Users can create SLURM job scripts to submit jobs to the system. These scripts can, and should, be modified in order to control several aspects of your job, like resource allocation, email notifications, or an output destination.


To submit a job script from one of the login nodes accessed via hpg.rc.ufl.edu, use the following command:

$ sbatch <your_job_script>

To check the status of submitted jobs, use the following command:

$ squeue -u <username>

View SLURM Commands for more useful SLURM commands.

Managing Cores and Memory

See Account and QOS limits under SLURM for the main documentation on efficient management of computational resources, and an extensive explanation of QOS and SLURM account use.

The amount of resources within an investment is calculated in NCU (Normalized Computing Units), which correspond to 1 CPU core and about 3.5GB of memory for each NCU purchased. CPUs (cores) and RAM are allocated to jobs independently as requested by your job script.

Your group's investment can run out of **cores** (SLURM may show QOSGrpCpuLimit in the reason a job is pending) OR **memory** (SLURM may show QOSGrpMemLimit in the reason a job is pending) depending on current use by running jobs. The majority of HiPerGator nodes have the same ratio of about 4 GB of RAM per core, which, after accounting for the operating system and system services, leaves about 3.5 GB usable for jobs; hence the ratio of 1 core and 3.5GB RAM per NCU.

Most HiPerGator nodes have 32 cores and 128 GB RAM (~30,000 cores in the newer part of the cluster) or 64 cores and 256 GB RAM (~16,000 cores in the older part of the cluster). The bigmem nodes and the newer Skylake nodes have a higher ratio of 16 GB/core and 6 GB/core, respectively. See Available_Node_Features for the exact data on resources available on all types of nodes on HiPerGator.

You must specify both the number of cores and the amount of RAM needed in the job script for SLURM with the --mem (total job memory) or --mem-per-cpu (per-core memory) options. Otherwise, the job will be assigned the default 600mb of memory.

If you need more than 128 GB of RAM, you can only run on the older nodes, which have 256 GB of RAM, or on the bigmem nodes, which have up to 1.5 TB of RAM.

Monitoring Your Workloads

You can see presently running workloads with the squeue command e.g.

$ squeuemine

OpenOnDemand offers a method to monitor jobs using the Jobs menu in the upper toolbar on your dashboard. This will show your current running, pending, and recently completed jobs. Select: Jobs -> Active Jobs from the upper dashboard menu.


We provide a number of helpful commands in the UFRC module. The ufrc module is loaded by default at login, but you can also load the ufrc module with the following command:

$ module load ufrc


Examples of commands for SLURM or HiPerGator specific UFRC environment module

Expand this section to view Examples of commands

$ slurmInfo           - displays resource usage for your group
$ slurmInfo -p        - displays resource usage per partition
$ showQos             - displays your available QoS
$ home_quota          - displays your /home quota
$ blue_quota          - displays your /blue quota
$ orange_quota        - displays your /orange quota
$ sacct               - displays job id and state of your recent workloads
$ nodeInfo            - displays partitions by node types, showing total RAM and other features
$ sinfo -p partition  - displays the status of nodes in a partition
$ jobhtop             - displays resource usage for jobs
$ jobnvtop            - displays resource usage for GPU jobs
$ which python        - displays path to the Python install of the environment modules you have loaded

Getting help

If you are having problems using the UFRC system, try our resources at Help or let our staff know by submitting a support request.