Skip to content

ed-aisys/Server-Info

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Cluster

Server Information

gala1

  • 4U server - PCIe 4.0 (Chassis similar to Gigabyte G482-Z54)
  • CPU: 2 x 28 CPU cores (AMD EPYC Zen 3, 7453)
  • GPU: 8 x Nvidia A5000 (NVLinked, 4 pairs)
  • Memory: 1TB DDR4 3200MHz (16 x 64GB)
  • NVMe SSD: 2 x 3.84TB (Intel P5510, PCIe 4.0)
  • SATA SSD: 1.92TB (Intel S4510)
  • 1 Gbps NIC (Future expansion to have InfiniBand)

G482-Z54

jazz1

  • Workstation server - PCIe 4.0
  • CPU: 1 x 16 CPU cores (AMD Ryzen 9, 5950X)
  • GPU: 1 x Nvidia 3090
  • Memory: 128 DDR4 3200MHz (4 x 32GB)
  • NVMe SSD: 1 x 2TB (Samsung 980 Pro, M.2, PCIe 4.0)
  • SATA SSD: 1 x 4TB (Samsung 870 QVO)

Jetson AGX Orin Developer Kit

  • CPU: 1 x 12 CPU cores (Arm® Cortex®-A78AE v8.2)
  • SD Card: 1 x 128GB
  • NVMe SSD: TBD

IMPORTANT: The server can only be accessed within Informatics firewall through Ethernet cables in side building or proxy jump through ssh.inf.ed.ac.uk

Create User

sudo useradd -m -s /usr/bin/bash -G sudo <username> # user WITH root privileges
sudo useradd -m -s /usr/bin/bash <username>         # user WITHOUT root privileges

sudo passwd <username>             # set user password

After the default passward is set for users, they can login and change their own password by passwd

Storage

sudo usermod -aG storage <username>

For non-sudoers, do the following instead

sudo mkdir /mnt/raid0nvme1/<username>
sudo chown <username>:<username> /mnt/raid0nvme1/<username>

IMPORTANT: Home directory is only for downloads, compilation. Put data on NVMe SSD under /mnt/raid0nvme1

Remote Access

For users want to have ssh access by public/private key pair

ssh-copy-id <username>@gala1   # execute on local machine with key pair
ssh <username>@gala1

Software Stack

Add user to anaconda group

sudo usermod -aG anaconda3 <username>
sudo usermod -aG docker <username>      # docker access
sudo usermod -aG microk8s <username>    # standalone k8s accress

Add the following to .bashrc or .zshrc

# >>> conda initialize >>>
# !! Contents within this block are managed by 'conda init' !!
__conda_setup="$('/home/gala1/anaconda3/bin/conda' 'shell.bash' 'hook' 2> /dev/null)"
if [ $? -eq 0 ]; then
    eval "$__conda_setup"
else
    if [ -f "/home/gala1/anaconda3/etc/profile.d/conda.sh" ]; then
        . "/home/gala1/anaconda3/etc/profile.d/conda.sh"
    else
        export PATH="/home/gala1/anaconda3/bin:$PATH"
    fi
fi
unset __conda_setup
# <<< conda initialize <<<

sudo usermod -aG gurobi <username>

export GUROBI_HOME="/home/gala1/gurobi951/linux64"
export PATH="${PATH}:${GUROBI_HOME}/bin"
export LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:${GUROBI_HOME}/lib"

CUDA

export CUDA_HOME=/usr/local/<cuda version>
export PATH=${CUDA_HOME}/bin:$PATH
export LD_LIBRARY_PATH=${CUDA_HOME}/lib64:$LD_LIBRARY_PATH

Disks

(Thu 31 Aug 2023 05:21:42 PM BST)

/dev/sda3       1.8T  1.2T  491G  71% /
/dev/md0        7.0T  4.7T  2.0T  72% /mnt/raid0nvme1
/dev/sdb1       3.6T  2.7T  737G  79% /mnt/data
/dev/md1         15T  4.7T  9.8T  33% /mnt/raid0sata1
...

Please do not download dataset directly to /home directory.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •