Articles

Sys admins are like smokejumpers who parachute into fires, fighting them until they are out, or at least under control. When you jump into the fire, you only have the tools you brought with you.

One way to share HPC systems among several users is to use a software tool called a resource manager. Slurm, probably the most common job scheduler in use today, is open source, scalable, and easy to install and customize.

The Python SymPy library for symbolic mathematics allows you to create complex mathematical expressions.

HPC systems require shared filesystems to function effectively. Two really good choices for both small and large systems are NFS and SSHFS.

The indispensable Lmod high-performance computing tool allows users to control their build and execution environment.

The pdsh parallel shell tool lets you run a command across multiple nodes in a cluster.

Building HPC applications for production systems is never easy, especially when containers are involved, but with Python and HPC Container Maker, you can describe the container you want quickly and easily without having to worry about the details.

Sometimes your Python programs need a little more speed. The pyamgx library can help you speed up your Python code.

Let users recover a deleted file without admin intervention by aliasing the rm command with mv or by writing your own script that moves the data to another location.

The Linux kernel has several I/O schedulers that can greatly influence performance. We take a quick look at I/O scheduler concepts and the options that exist within Linux.