Storage Details

Storage Tiers & Configuration

The TU HPC storage infrastructure is divided into multiple storage tiers, each designed for specific use cases:

1. Software Storage

• Location: /dev/sdb1
• Capacity: 470GB
• Purpose: Storing installed software, modules, and system libraries.
• Access: Read access for all users, write access for administrators.

2. Scratch Storage (High-Speed Temporary Storage)

• Location: /mnt/storage0
• Capacity: 1.9TB
• Purpose: High-speed temporary storage for running computations.
• Policy:
• Non-persistent: Data may be deleted after job completion.
• Users must regularly move important data to home or project storage.

3. Home Directory (User Personal Storage)

• Location: /home
• Capacity: 1.8TB
• Purpose: User-specific data, research scripts, personal files.
• Quota: Users have a designated storage limit to ensure fair use.
• Backup: Take your own backup.

4. Memory File System (Temporary In-Memory Storage)

• Location: tmpfs
• Capacity: 126GB
• Purpose: High-speed memory-based temporary storage.
• Use Cases: Best for temporary caching and intermediate data processing.

Data Management Policies

Storage Usage Guidelines

• Scratch Storage (/mnt/storage0) should be used for active computations only, which can be also assessed inside (/home/<user-name>/storage0).
• Long-term storage should be maintained in the home directory (/home).
• Users must monitor their storage usage and clean up unnecessary files.

Backup & Data Security

• Home directories are not regularly backed up.
• Scratch storage is NOT backed up – users should save important data elsewhere.
• Users are responsible for managing their research data securely.

Access & Permissions

• Software storage (/software) is managed by administrators.
• User files in /home are private but can be shared with appropriate permissions.
• Group storage options available for collaborative research projects.

Future Storage Expansion

TU HPC is continuously upgrading its storage infrastructure to support:

• Larger storage volumes for big data research.
• Faster access speeds for computationally intensive workloads.
• Enhanced data redundancy and recovery mechanisms.

For details on managing your data efficiently, visit the Data Management Guide.

Optimized storage solutions for high-performance computing! 🚀