Skip to content
This repository was archived by the owner on Mar 23, 2026. It is now read-only.

UAMLDA-Resources.md

Latest commit

 

History

History
78 lines (55 loc) · 3.38 KB

UAMLDA-Resources.md

File metadata and controls

78 lines (55 loc) · 3.38 KB

About the UAMLDA Computing Resources

All of the computing facilities in the UAMLDA given a codename, and these codenames are characters from the Game of Thrones series. There is currently only one server in the group which is described below. Any group member can ssh into the server; however, you must be connected to the Univeristy of Arizona's VPN if you are off campus.

The Research Server

This a private server to the students working in the University of Arizona's Machine Learning and Data Analytics Lab. The specifications are

  • Intel Xeon E7-4830 v3 Dodeca-core (12 Core) 2.10 GHz
  • RAM 16 SM 16GB DDR3 SDRAM RAM 16 GB (1 x 16 GB) DDR3 SDRAM-1600 MHz
  • Management 1 Integrated IPMI 2.0 & KVM with Dedicated LAN
  • Controller 1 Dual-Port Intel i350 GigE plus 2 Ports 6Gb/s SATA and 4 Ports 3Gb/s
  • RAID controllers 1 LSI 9361-8i (8-Port Int, PCIe 3.0, 1GB Cache, Dual Core)
  • RAID controllers 1 LSI 9361 KIT CACHEVAULT
  • Hot-Swap Drive 5 Seagate Constellation 3 TB 3.5" Internal Hard Drive - SAS - 7200 rpm
  • Power Supply 1 Redundant 1400W Power Supply with DSC and PMBus
  • Management SW 1 IPMI Support for Intelligent Platform Management Interface v.2.0

Configuring and Accesing the Server

ssh Configuration

If you're lazy you can make the ssh process a bit easier by adding the following to the ~/.ssh/config file on your local machine.

Host khaleesi 
  User UAnetID
  Hostname <server name>.ece.arizona.edu

To make your life even easier, add your ssh key to the server.

# do not add the <>
$ ssh-keygen -t rsa -b 4096 -C "your_email@example.com"
$ ssh-add ~/.ssh/id_rsa
$ ssh-copy-id <UAnetID>@<server name>.ece.arizona.edu

Setting Up Your Paths

Here is an example, however, you will need to make the appropriate changes for your Anaconda installation.

. /home/skel/bash/bashrc

export PATH=~/bin:${PATH}
export PATH=/usr/local/MATLAB/R2015b/bin/:${PATH}
export PATH=/scratch/ditzler/anaconda2/bin/:${PATH}

Accessing RStudio

The ssh accounts are dealt with via the active directories that are mounted from the University of Arizona; however, the RStudio accounts are managed locally on the servers. The usernames for Rstudio are rs_netid (e.g., rs_ditzler). The default path for the Rstudio accounts are /localhome/rs_netid. Note that your Rstudio and standard ssh accounts are different and cannot read/write to both accounts. I recommend using Git to manage your projects in the localhome directory.

The Rstudio accounts can be accessed by http://<server name>.ece.arizona.edu:8787.

Setting up Python

The Anaconda distribution of Python is installed on the server; however, there are a lot of permission issues that arise when new modules are install. Therefore, to keep things simple, everyone should have their own version of Anaconda installed. You will therefore be required to keep up with the latest updates and package management.

To install a vanilla version of Anaconda, run:

wget https://3230d63b5fc54e62148e-c95ac804525aac4b6dba79b00b39d1d3.ssl.cf1.rackcdn.com/Anaconda2-2.4.1-Linux-x86_64.sh
bash Anaconda2-2.4.1-Linux-x86_64.sh
# follow the instructions and install to your /scratch/ folder

Extra packages I installed are

conda install -c https://conda.anaconda.org/jjhelmus tensorflow
conda install cvxopt
conda install opencv
pip install dicom
pip install lmdb