Home

This is a server and remote experiment workflow guide for Visual Information Science Laboratory members in Hiroshima University. It covers the basic path from preparing your local computer, SSH login, key setup, VS Code remote development, and file transfer to long-running experiments with tmux and shared-resource guidelines.

PBRT-v4 spectral GPU path tracing render of the San Miguel de Allende courtyard scene. Rendered with the volpath integrator at 4096 samples per pixel on an NVIDIA RTX PRO 6000 Blackwell Max-Q GPU. Final homepage crop: 3840 x 2160. Total render time: 5 min 39 sec.

Before You Connect

Prepare Environment Local Environment

Choose a suitable terminal, confirm that SSH is available, and understand the difference between the lab Zone C network, campus VPN, and off-campus networks.

Connect to Servers

First Login SSH and Server Login

Learn how to log in to a server with ssh username@server-address.

Recommended Setup SSH Public and Private Keys

Generate an Ed25519 key, add the public key to the server's authorized_keys, and set local aliases such as vis-server in SSH config.

Off-campus Network Off-campus Access

Understand the difference between the regular campus SSL-VPN and Zone C VPN.

Daily Remote Work

Terminal Basics Common Linux Commands

Collects common commands for server jobs and everyday terminal work.

Remote Development VS Code Remote SSH

Use VS Code to connect to a remote server, edit code, and manage project files.

Data Transfer File Transfer

Choose VS Code, scp, rsync, sftp, or a graphical SFTP tool based on file size and transfer scenario.

Running Experiments

Experiment Environments Python Environment Isolation

Install Miniforge and use a separate conda environment for each project to avoid dependency conflicts.

Long-running Tasks tmux and Running Experiments

Run long experiments safely.

Shared Resources GPU / Disk / Memory Guidelines

Before starting training, check nvidia-smi, free -h, df -h, and directory usage to avoid putting unnecessary pressure on server resources.

Troubleshooting and Site Information

Q&A FAQ

This page will collect common issues with connections, permissions, environments, and experiment runs.

Site Analytics Analytics

View the public analytics dashboard, including visits, referrers, countries/regions, devices, browsers, and popular pages.

Security Reminder

Do not disclose server passwords or SSH private keys. Do not expose lab server ports, configure reverse tunnels, use third-party mesh networking tools, or bypass network access methods approved by the university and lab without permission.

GitHub contributors

People whose commits keep this guide useful for the lab.

View all