Starting on July 1, 2016, the system is now open to the XSEDE community! You will need a XSEDE account and an awarded allocation to get access to XStream. Please discover XSEDE resources and services through this Getting Started Guide.
The XStream User Guide for XSEDE users is available here.
To request an account, the sponsoring Stanford faculty member participating on the NSF MRI grant should email email@example.com specifying the names and SUNet IDs (usernames) of his/her research team members needing an account.
XSEDE users must use the SSO Hub.
For Stanford users, XStream requires Kerberos for authentication security. Kerberos software needs to be installed and configured for Stanford on your client. If you don’t have Kerberos already setup, please see the Kerberos setup page.
On Linux or Mac, to get a valid Stanford Kerberos ticket, use the following command and enter your SUNet ID password:
Your Kerberos ticket is valid for 25h. At any time, you may check that it is still valid using the
To login to XStream, you need to have an SSH client on your local machine.
You will use the
sftp tools to move data to and from XStream. Note that in the near future, you will also
be able to use Globus tools for data transfer.
To actually login to XStream, you need a valid Kerberos ticket (see above) and tell ssh to use Kerberos. To do so,
please first add the following lines to your
$HOME/.ssh/config (create it if needed):
and then use:
Or, you can specify the SSH GSSAPI options directly in command line using
In either way, you shouldn’t be prompted for a password.
It is recommended to use one of the following round-robin DNS aliases to reach a login node:
If you have a
screen running on a specific login node, you can directly use the following hostnames:
Please, do NOT use the login nodes for computationally intensive processes. These nodes are meant for compilation, file editing, simple data analysis, and other tasks that use minimal compute resources. All computationally demanding jobs should be submitted and run through the batch queuing system. You may however use the few GPUs available on the login nodes to perform simple and short tests. Please note that GPUs on the login nodes are configured in Default Compute Mode, meaning that multiple contexts are allowed per device. They are not suitable for performance evaluation.
Instead, submit batch jobs or request interactive access to the compute nodes as detailed in the Slurm page.