Actions

Difference between revisions of "Windows stories"

From ALICE Documentation

(Windows-Getting started)
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
==[[Windows-Getting started]]==
+
{{:Windows-Getting started}}
  
Introduction to HPC
+
{{:Windows-Advanced guide}}
Getting an ALICE account
 
Connecting to the HPC infrastructure
 
Running batch jobs
 
Running interactive jobs
 
Running jobs with input/output data
 
Multicore jobs/Parallel Computing
 
[[Troubleshooting]]
 
ALICE policies
 
Frequently Asked Questions
 
  
==[[Windows-Advanced guide]]==
+
{{:Software-specific Best Practices}}
Fine-tuning Job Specifications
 
Multi-job submission
 
Compiling and testing your software on the HPC
 
Checkpointing
 
Program examples
 
Job script examples
 
Best Practices
 
Graphical applications with VNC
 
Graphical applications with X2Go
 
HPC-UGent GPGPU cluster
 
 
 
==[[Software-specific Best Practices]]==
 
MATLAB
 
OpenFOAM
 
Mympirun
 
Singularity
 
SCOOP
 
Easybuild
 
Hanythingondemand (HOD)
 
 
 
SLURM options
 
Useful Linux Commands
 

Latest revision as of 12:11, 14 June 2020

Best Practices

Know when you are on a login node.  You can use your Linux prompt or the command hostname.  This will tell you the name of the login node that you are currently on. Note that the ssh gateway host itself is a secure portal from the outside and serves no compute function.

  • Appropriate activities on the login nodes:
    1. Compile code, Developing applications,
    2. Defining and submitting your job,
    3. Post-processing and managing data,
    4. Monitoring running applications.
    5. Change your user password.
  • Avoid computationally intensive activity on the login nodes.
    1. Don't run research applications.  Use an interactive session if running a job is not appropriate.
    2. Don't launch too many simultaneous processes.  While it is fine to compile on a login node, avoid using all of the resources. For example "make -j 14" will use half of the cores.
    3. That script you run to monitor job status several times a second should probably run every few minutes.
    4. I/O activity can slow the login node for everyone, like multiple copies or "ls -l" on directories with 000's of files.
  • Hyperthreading is turned off.  Running multiple threads per core is generally not productive.  MKL is an exception to that if it is relevant to you.

Getting a User account for ALICE

To get a user account on ALICE, mail a request to helpdesk@alice.leidenuniv.nl (mail request) and provide the following information

  • your name
  • your ULCN or LUMC account name
  • your university or institute e-mail address.
  • What do you plan to do on ALICE (1 or 2 sentences are sufficient)?
  • Do you plan to process sensitive data (related to persons) or other data that must be treated confidentially?


At ALICE, we will create local ALICE accounts with the same account name.

If you are a student, we need confirmation from your supervisor that you require access to ALICE.

Once we have created your account, you will receive an e-mail from admin@alice.leidenuniv.nl providing you with an initial password to log in

IMPORTANT: Make sure to change the password in our e-mail when you log in to ALICE for the first time. You can use the command passwd for this.

IMPORTANT: Please read the section on Acceptable Use before requesting an account.


Acceptable use policy

Complience with University policy

High Performance Computing (HPC) facility users are responsible for complying with all University policies, including Acceptable Use of Computers.

Resource policy

The supercomputers represent a unique resource for the campus community. These computers have special characteristics that are not found, or are of limited availability, on other central computers, including parallel processing, large memory, and a Linux operating system.  The allocation of High Performance Computing (HPC) resources requires close supervision by those charged with management of these resources.  

The login nodes are designated for small, short interactive jobs, and submitting batch jobs and not for running compute jobs.

Data Storage Policy

ALICE does not provide support for any type of controlled data. No controlled data (GDPR, HIPAA, EAR, FERPA, PII, CUI, ITAR, etc.) can be analysed or stored on any HPC storage. Users must not transfer sensitive data (data related to people) to ALICE. Data must be anonymized before it can be transferred to ALICE. In case you are unsure about the contents/classification of the data, please contact the helpdesk.

ALICE is not a datamanagement system where research data can be stored for longer periods of time. All data that is transferred to ALICE must be copies of data. Users must make sure that data that is transferred to ALICE remains available somewhere else. All data with value that is generated on ALICE must be moved off ALICE as soon as this is possible after the job completed.

Data in the user’s home directory is backed up (see Backup & Restore). The home directory is intended to store scripts, software, executables etc, but is not meant to store large or temporary data sets.

Login to ALICE from Windows

There are multiple ways to connect to ALICE from Windows. Below a list in order of complexity using:

Powershell

MobaXterm

Putty

WSL

Windows-Advanced guide

Fine-tuning Job Specifications

Multi-job submission

Compiling and testing your software on the HPC

Checkpointing

Program examples

Job script examples

Best Practices

Graphical applications with VNC

Graphical applications with X2Go

ALICE GPU cluster

Software-specific Best Practices

MATLAB OpenFOAM Mympirun Singularity SCOOP Easybuild Hanythingondemand (HOD)