Getting Started

Server Resources

There are two kinds of computing services for running jobs that ARC provides, Compute Servers and Clusters.

Compute Servers

When you make a research account you will have access to our compute servers Circe and Hecate by default. These servers are large computers running linux that allow multiple users to run jobs simultaneously. Once you are connected to a compute servers, you can run jobs in the command line like you would on a normal linux computer, just be mindful of other users and their jobs.

Compute Clusters

A compute cluster is a set of networked computers that act as one. Compute clusters provide much more power for parallel jobs than the compute servers Circe and Hecate. We have two clusters here at ARC, Hydra and Gravel. If you want access to the compute clusters you will have to request access in addition to your regular research account. To request access to ARC’s compute clusters please contact: consultants@pdx.edu

Connecting to Our Servers

ssh is used to connect to our servers. ssh is included on Mac OSX and Linux and can used via the command line. If you are on Windows you will need additional software, most commonly PuTTY, to access our servers over ssh.

Connecting From Off Campus

If you are off the campus network you will have to use PSU’s vpn client OR ssh onto hera.rc.pdx.edu and then ssh onto the server you want to access from there.

Additional Information About Connecting

Storage

There are several options for storage on our systems.

Home Directories

Your home directory (~) is your own personal default storage space. When you connect to a server via ssh you are in your home directory. Home directory space is very limited. Do not store large files here–only configuration files, documents, and locally built programs. The compute servers share home directories.

Scratch Drives

Each server has its own Scratch Drive or scratch space. Scratch Drives are fast storage devices that you can use to store the data you want to run your jobs on. Scratch Drives are only intended to be used as short-term storage so please move your data off of scratch when you are done with your analysis.

If you are going to use a scratch drive please make a directory named after yourself to keep all of your files in.

Scratch drive locations

Compute Servers (Circe/Hecate):
  1. /disk/scratch
Hydra (compute cluster):
  1. /scratch
  2. /scratch2

Note

Scratch Space is not shared between the servers. e.g: scratch on Hecate is separate from scratch on Circe

Research Shares (I-Drive)

Research Shares are shared folders that you need to request access to. To request access to a share OR request a separate research share for you project please go to: http://intranet.pdx.edu/oit/i-drive-request

Research shares can be found in: /vol/share on all systems (excluding gravel)