Frequently Asked Questions

Can I run containers?

While we do not support docker, you can run containers using singularity. To use singularity run module load singularity.

$ module load singularity
$ singularity pull library://sylabsed/linux/alpine
WARNING: unable to verify container: alpine_latest.sif
WARNING: Skipping container verification
$ singularity shell alpine_latest.sif
Singularity> cat /etc/os-release
NAME="Alpine Linux"
ID=alpine
VERSION_ID=3.8.1
PRETTY_NAME="Alpine Linux v3.8"
HOME_URL="http://alpinelinux.org"
BUG_REPORT_URL="http://bugs.alpinelinux.org"
Singularity>

Please note that building the container image is not supported. You must build the image outside of the cluster on a system in which you have root access.

How do I run Python based frameworks (PyTorch, etc.)?

We’ve found that it’s best if users manage their own versions of Python and requisite dependencies. To make this easier, we have a current version the anaconda installer available. Simply run the installer and follow the prompts to install this for your user account.

$ /export/software/anaconda/install-current.sh

Welcome to Anaconda3 2020.07

In order to continue the installation process, please review the license
agreement.
Please, press ENTER to continue
>>>

How can I get CPU and memory stats from the node my job is running on?

First, determine the node that your job is running on

$ qstat -xf <job_id> | grep exec_host
exec_host = vsl084/1*64

Now, set up a port forward to the cluster to get access to the grafana dashboard.

ssh -L 3000:iam-adr-web:3000 ssh-iam.intel-research.net

You should now be able to point your browser to http://localhost:3000 and see a cluster usage dashboard. Go to http://localhost:3000/d/wk1o4drWk/node-dashboards?orgId=2 and choose the node you identified above to see detailed node information about the node your job was running on.