Getting Help

ACE Uganda support staff are here to help you with account issues, software questions, job failures, and general HPC guidance. Before submitting a support request, take a moment to review the tips below — a well-written request helps us resolve your issue faster.

How to Contact Us

Channel	Details
Email	support@ace-bioinformatics.org
Hours	Monday – Friday, 9:00 AM – 5:00 PM EAT (excluding public holidays)
Response time	Typically within 1–2 business days

Before You Submit a Request

Do some initial troubleshooting on your own — it's often the fastest path to a solution:

• Check the documentation. Browse the FAQs, Good Conduct guidelines, and Tutorials sections of this wiki.
• Read the error message carefully. Most error messages describe exactly what went wrong. Search the message text online if it's unfamiliar.
• Check your job output files. Review .out and .err files from your job for clues. Run sacct -j <jobid> --format=JobID,State,ExitCode,MaxRSS for resource usage details.
• Verify your environment. Run module list to confirm the right modules are loaded, and check that file paths in your scripts are correct.

Writing an Effective Support Request

A clear, detailed request allows us to diagnose your issue without multiple back-and-forth emails. Include the following:

What to include

1. Your username on the cluster.
2. What you were trying to do — a brief description of your goal.
3. What you did — the exact commands or steps you ran.
4. What happened — paste the complete, verbatim error message or unexpected output. Do not paraphrase errors.
5. What you expected to happen instead.
6. Job details (if applicable):
   • Job ID(s): find with sacct -u $USER
   • The directory containing your job script and data
   • Your job script (attach or paste it in full)
   • Any loaded modules (module list)
7. What you already tried to fix the issue.

Example of a good support request

Subject: Job 485210 — out-of-memory error running BLAST

Hi, I'm user jdoe. I submitted a BLAST job (ID 485210) in /home/jdoe/blast_project/ using the script run_blast.sh. The job was killed after 20 minutes with the error:

slurmstepd: error: Detected 1 oom_kill event in StepId=485210.0. Some of the step tasks have been OOM Killed.

I requested 4 GB of memory with --mem=4G. My input database is 12 GB. I suspect I need more memory but I'm not sure how much to request. I tried 8 GB and it still failed. Could you advise?

Example of a poor support request

Subject: Job not working

My job failed. Can you help?

The more context you provide upfront, the faster we can help.

What We Can Help With

• Account issues — password resets, access problems, quota increases
• Job failures — debugging Slurm errors, resource allocation issues, out-of-memory kills
• Software — questions about installed software, module loading, and custom installations
• Data management — file transfers, storage quotas, data organization
• General HPC guidance — choosing the right resources for your workload, parallelization strategies

What Falls Outside Our Scope

Our team focuses on cluster operations and general HPC guidance. Some requests are better addressed elsewhere:

• Debugging your application code — we can help with HPC-specific issues (MPI errors, Slurm failures), but we cannot debug the logic of your research code.
• Teaching programming languages — we recommend online resources, workshops, or your department's training programs for learning Python, R, etc.
• Third-party software internals — we can help install and configure software, but for application-specific scientific questions, consult the software's own documentation or community forums.

Stay Informed

• System announcements — maintenance windows, outages, and updates are communicated via email to all registered users. Make sure your contact email is up to date.
• Documentation updates — this wiki is regularly updated. Check back for new tutorials and guides.