4. Condor quickstart

A.K.A Condor for Busy People!

4.1. prerequisites

  1. Read Introduction to HTCondor for a general overview of the Surrey GPU facility.

  2. Ensure you have a valid Linux account for Accessing Condor (required to log into any university managed Linux machine).

Important

If you are in FEPS you will have this by default. If you are from another faculty you will need to Open a support ticket requesting Linux access.

  1. Request a project space to store your code and models in if you don’t already have one.

4.2. Connect to a Condor pool

The AI @ Surrey cluster offers most resources, and is available to any University member. To use it you must pass the corresponding SurreyLearn module. See AI @ Surrey for details.

The course is designed to give you the basic skills you will need to use HTCondor and more specifically the AI@Surrey cluster. It covers an introduction to the network and cluster, how to run your jobs and example jobs you can adapt to get you started.

See this page for details, including the access requirements, for all the available Condor pools at Surrey.

Once you have identified the Condor pool you want to use. Log into the submit node via SSH:

Connecting via SSH to the aisurrey-condor submit node
$ ssh abc123@aisurrey-condor.surrey.ac.uk

You will need to substitute your username and also the hostname of the Condor submit node you want to connect to.

You can now start Submitting jobs

4.3. Further reading

Get involved…

Join our Research Computing Community on Microsoft Teams