These servers are dedicated to the tools used by astrophysicists for data processing and numerical simulation.
Through a queue management system, the cluster allows to launch processes on many processors by optimizing the use of all the computation nodes.
It is composed of two calculation queues:
– A queue dedicated to massively parallel computations
– A queue dedicated to single-core batch computations or limited to a server
The Redmine farm allows to quickly implement a knowledge exchange system for the LAM teams and projects. (https://projects.lam.fr/)
Computing project environment
Using these tools, researchers and engineers have access to a software version management system, bug tracking and enhancements usable via web interfaces.
– Redmine project management server https://projects.lam.fr
– Application development management server such as gitlab http://gitlab.lam.fr
– Continuous compilation and code analysis server
Information systems environment
By duplicating servers and virtualizing applications to allow high fault tolerance, this platform allows for the continuous availability of scientific data.
With a server virtualization system and an application virtualization system, we can set up virtual machines for the development and provision of data and tools in a way that is transparent to the laboratory’s development and users.