Good Conduct on ACE HPC
As a shared resource, ACE HPC serves many researchers simultaneously. Your actions directly impact the experience of other users. Following these guidelines ensures equitable access and optimal performance for everyone.
Core Principles
- Do not run jobs on login nodes - Login nodes are shared gateways, not compute resources
- Respect file system quotas - Exceeding quotas affects your jobs and strains shared storage
- Submit jobs responsibly - Request only the resources you need
- Transfer files efficiently - Use appropriate methods to avoid overloading the network
Quick Reference
| Do | Don't |
|---|---|
| Submit jobs via SLURM | Run computations on login nodes |
| Monitor your quota usage | Exceed allocated storage |
Organise job output in $HOME/jobs/ | Leave large result files in $HOME indefinitely |
| Archive small files before transfer | Transfer thousands of tiny files |
| Request appropriate resources | Over-request nodes or time |
Why This Matters
The ACE cluster is a shared community resource. When one user monopolizes login nodes or overwhelms the file system, it degrades performance for everyone. Responsible conduct ensures:
- Fair access - All researchers can work effectively
- System stability - Infrastructure remains healthy and responsive
- Faster job turnaround - Resources are used efficiently
Consequences of Misuse
Repeated violations of good conduct guidelines may result in:
- Account suspension
- Reduced resource allocations
- Required consultation with support staff before reinstatement
Getting Help
If you're unsure whether your workflow follows good conduct guidelines, contact the ACE support team before running large jobs. We're here to help you succeed while maintaining a healthy shared environment.
- Email: support@ace-bioinformatics.org
- See the FAQ for common questions