HPCT

High Performance Computing Tools

The learning objectives of this course unit are:

Integrate cluster hardware consisting of several computing and storage nodes into a supercomputer
Configure the system services that enable efficient management of the cluster’s hardware and software
Installing software and supplying it to various users
Compile end-user applications and run them on various nodes
Analysing system and application performance
Formulate security policies and apply tools to reinforce the system, such as firewalls and intrusion detection
Describe and document system configuration

The course covers several important aspects of high-performance tools, including:

Introduction to LINUX : Administration Basics, File System, SSH, Scripts, Environment Variables
High-performance infrastructure architectures in particular Interconnects and network topologies in high-performance infrastructures
Task management as SLURM
Administrative task automation tools such as Ansible
Containers in high-performance infrastructures such as Singularity/Apptainer
Compiling tools such as Easy Build and Spack
Cluster management tools such as Warewulf and IPMI
Security in high-performance infrastructures: firewalls, certificates and PKI, security policies
Tools for monitoring and diagnosing the system such as Kibana/InfluxDB, Grafana

High-Performance Computing: Modern Systems and Practices. Thomas Sterling, Matthew Anderson, Maciej Brodowicz. 2017. Morgan Kaufmann
Modern System Administration: Managing Reliable and Sustainable Systems. Jennifer Davis. 2022 O’ Reilly Media, Inc
Ansible: Up and Running. Bas Meijer, Lorin Hochstein, René Moser. 2022. O’ Reilly Media, Inc