Most Common Issues

Segmentation fault

This error occurs when your code tries to access memory outside its allocated range. Sometimes, increasing the memory allocation can resolve this issue.

Slurmstepd error

Indicates that the job has exceeded the memory limit. The default memory per core is 4 GB. Increasing memory using the Slurm parameters --mem or --mem-per-cpu usually resolves this problem.

Requesting excessive resources

If you request more resources than are available, your job will remain in the queue. Test your code first to determine its actual CPU and memory requirements, and request only what you need.

Over-reserving resources

Requesting far more resources than necessary can cause delays for other users by keeping their jobs queued unnecessarily.

User-installed software issues

Prefer using pre-installed modules on the cluster before installing or compiling your own software.

Anaconda virtual environments

These can be complex to configure. If you encounter errors, search for solutions online; many users have shared fixes for similar problems.

For any issues or assistance, please contact it@ku.edu.tr.