Frequently Asked Questions

[edit] General

How do I submit a job on Athena?

Please read theJob Submission Tutorial and the section on Job Routing and Priority Management.

Can I run jobs on the head node?

NO! The head node is strictly for compiling your code and submitting jobs. If you start running things on this nodes, you will mess up everyone else's job on the entire cluster. Remember, Athena is a shared resource. Be polite.

Where is MPI?

See the section on Using MPI

Where do I install software?

At the moment, you can install software in your home directory. We are looking into supporting a community software area.

Can I request software be installed?

Yes. Send email to help@phys.

[edit] Questions regarding Athena downtimes

Last time a security breach was detected, Athena was down all week! I mean, geez, how long does it take to re-install the operating system?!!

The support team usually takes the opportunity to conduct a substantial number of needed upgrades during any re-install. Due to the small number of staff we have, it is important that every opportunity is taken to combine work.

If the operating system is re-installed, is my data on Athena going to be wiped?

This drastic action would only be taken if there was a preponderance of evidence to suggest that the user data partitions presented a security threat. You will be notified beforehand if such evidence exists.

But are you really sure that my data won't be wiped due to the upgrade/reinstall process?

Actually, whenever a node on Athena is rebooted due to an anomalous failure, it is automatically rebuit from scratch. So this kind of thing happens all the time without consequence.

Why weren't these upgrades be done more regularly so that the lengths of downtimes can be shorter?

In order to support jobs that run on the order of one week, it is difficult to bring the cluster down regularly.