Page tree
Skip to end of metadata
Go to start of metadata

Version 2018-05-16.

For any questions, contact antti.karttunen@aalto.fi. 

Introduction

  • Here you can find the instructions to use the Wihuri computing cluster on a computational chemistry course
  • If you want to use Wihuri for research purposes, please see Computing Resources at Aalto CMAT.  

Rules for using Wihuri computing cluster

  1. For the exercise sessions and related coursework, you will get a personal user account
  2. You are only allowed to use your account until the end of the course
  3. You are only allowed to carry out course-related work on the cluster
  4. Giving the details of the user account to anyone else is strictly forbidden
  5. Trying to install any software on the cluster is strictly forbidden
  6. Trying to modify any files outside the home directory of your user account is strictly forbidden

Using Wihuri

The address of the cluster is wihuri.pub.chemistrylab.aalto.fi. You can connect the cluster using SSH.

  • The cluster can only be connected from Aalto network or by using Aalto VPN (first start VPN, then connect with SSH)
  • To use the computing cluster, you need an SSH client. For Windows, PuTTY is an excellent (and free) choice. Just download putty.exe to any folder and execute it. No installation is necessary.
  • You will need a basic knowledge of Linux to use the cluster (good tutorials are available)
  • To transfer files between a cluster and your workstation, you can use SFTP. WinSCP is a reasonably good SFTP-client.
  • Any text files on the cluster can be edited over the SSH terminal connection by using a text editor. nano is an easy-to-use editor with a menu system. vi is a very convenient and fast editor with a bit steeper learning curve (a short vi reference might help). 
  • To edit a file, just execute nano myfile.txt or vi myfile.txt.

Submitting jobs

All calculations need to be run using a queuing system. Here is how it works:

  • Let us assume that in the current working directory there is an ORCA input file h2o.inp
  • The job can be submitted to the queue with the command jsub orca h2o.inp
  • You can check the status of the queue with the command jstat
  • You can delete a job from the queue with the command jdel JOB_ID (JOB_ID comes from jstat)
  • You can find the output from h2o.oout
  • If your job runs into trouble, please see the output file and h2o.batch-log for error messages

Important resource limits

On a computational chemistry course, you are allowed to run jobs using up to

  • 12 processors
  • 5 GB memory per processor

Do not exceed these limits unless the teacher tells you to do so.


  • No labels