Logging on to rhino for the first time

Updated: July 20, 2020

Edit this Page via GitHub       Comment by Filing an Issue      Have Questions? Ask them here.

This tutorial is a step-by-step introduction to logging on to rhino, a remote computing resource available to researchers at Fred Hutch. It assumes you have basic familiarity with navigating the command line using Unix commands like cd, ls, and pwd, and are connecting via a shell or equivalent emulator: on Mac, this would be Terminal; on Windows, this would be something like WSL or Git bash, a part of Git for Windows.

The basic process below below includes:

  • Logging on to snail (one option for connecting remotely)
  • Logging on to rhino
  • Grabbing a node to run an interactive job
  • Loading software
  • Exiting out of everything

This tutorial uses snail as the method for accessing the Fred Hutch campus network. This step is unnecessary if you are already connected to the network (such as through VPN). For more information about remote login, please see Access via a remote login.

Command prompts

We’ll be referencing your command prompt for the duration of this tutorial, which refers to the information printed at the start of each line in your shell where you enter commands. Command prompts generally take the form of:


The specific symbol at the end may be different for you on your own computer, but the shared compute resources we’ll be describing in this tutorial all use $. When you are logged on to Fred Hutch compute resources, username will always be your HutchNetID.

Logging on within the Fred Hutch network

This section assumes you are logging on from campus (e.g., connected to the Fred Hutch network). If you are logging in from off campus, please ensure you are connected to the Fred Hutch network using one of the options described in off campus log-in.

In your shell, type the following (where username is your HutchNetID):

ssh username@rhino

Hit enter to execute the command. The first time you execute this on a computer, you’ll receive a response similar to:

The authenticity of host 'rhino.fhcrc.org' can't be established.
Are you sure you want to continue connecting (yes/no)?

Type `y` and press enter.

You’ll be prompted to enter your username and then password (note that your cursor will not move as you type your password; this is normal!).

After you have successfully logged in, you’ll see the information about the cluster printed to your screen. You’ll be ready to start entering commands when you see a prompt like the following appear:


Logging on outside of the Fred Hutch network

Logging in to rhino as described above assumes you are already connected to the Fred Hutch network. If you are off campus, you will need to first connect to the campus network. Here we describe the use of snail for remote login, since it doesn’t require the use of additional software beyond what it already included in the tutorial. VPN may be a preferrable option for you, especially if you are already using it to access other Hutch resources. For more information about remote login, please see Access via a remote login.

In your shell, type the following (username is your HutchNetID):

ssh username@snail.fhcrc.org

You’ll see a message printed to the screen that starts with:

Welcome to the Fred Hutchinson Cancer Research Center

Your command prompt will now look something like:

[username@ssh-gw1-prd ~]$

Because you’ve already logged on to snail with your username, you can use the following shortcut to log on to rhino:

ssh rhino

Follow the process as described above to enter your password and access rhino.

Logging on to gizmo via grabnode

When you log on to the cluster, all commands you run are limited by the compute resources associated with your PI. (and by the cluster being shared by many people!). It’s good etiquette, and often a necessity, to request your own node for additional tasks.

You can accomplish this by executing the following command:


You’ll now be asked a series of questions about what compute resources you would like to request. We’ll describe requesting the smallest unit of resources here for demonstration purposes. When prompted:

  • type 1 for number of CPUs
  • hit Enter to select the default for memory
  • type 1 for number of days

A few extra pieces of information for new users:

  • When requesting resources for your own work, choose appropriate selections to suit the needs of the task. If you aren’t sure you need more than one CPU, you almost certainly do not.
  • There is not currently a way to execute grabnode with flags; you are required to use the interactive questions.
  • If you get an error like “Invalid Account” it indicates that your association to a PI account has not been completely set up- contact Scientific Computing.

You’ll note your command prompt will switch to say gizmo instead of rhino:


You can interact with your directories the same as you would normally, but now your commands will run on a part of the cluster allocated to you specifically. See this documentation for more information on when to use grabnode.

Loading software

Your compute environment is the collection of machinery, software, and networks on which you perform tasks. Managing environments is challenging, but there are tools available to help you stay on the right track.

rhino has hundreds of pieces of software installed, so it uses a module system to make software available for use. To see what software you currently have loaded:

module list

If you’re looking for a specific package, you can search using terms related to the name of the software. We’ll use FastQC as an example software:

module spider FastQC

If you see the package you want, you can load it with:

module load FastQC

…or the abbreviated version, which is the same as above:

ml FastQC

By default, the most recent version of the software is loaded. You can load specific versions by specifying them in the command:

ml FastQC/0.11.2

For this example, you should see in the output that there is also a module available to load as fastqc, and additional information on using the modules is included as well.

For more information on using the module system, please see the Computing Environments page, and you can view lists of available software on the Scientific Software page.

Logging out

Once you are done working on the cluster,
you should log out.

If you still have a command running, logging out will stop the process. See Moving beyond the basics for available solutions to this limitation.

The process for logging out depends how you’ve logged on when starting to work. In general, you’ll need to type exit to end each session you started. You can tell what session using the information in the command prompt.

For example, if you are logged on to an interactive job in gizmo:


Typing exit will end the interactive job and return you to rhino, at which point your command prompt will read:


Type exit again to end your session on rhino.

If you logged on using snail, you’ll see something similar to:

[username@ssh-gw1-prd ~]$

You’ll need to exit once more to leave the Fred Hutch network and see your command prompt return to your local computer.

Moving beyond the basics

This demo covers the basic process of logging on and using resources for the first time. Once you are comfortable with the basic process, you should look at the complete documentation on Large Scale Computing for information on more advanced tasks, like:

  • executing batch jobs using slurm, including limits and recommendations for creating scripts
  • using screen or tmux to keep processes executed on the command line, even after you’ve logged off
  • running applications like RStudio

Updated: July 20, 2020

Edit this Page via GitHub       Comment by Filing an Issue      Have Questions? Ask them here.