Services

In FCSCL different services are provided for using the infrastructure available to researchers at the center of public and private institutions and companies to develop R&D&i projects or those belonging to the field of technological innovation.

Calculation

For the execution of jobs in different clusters of the system, jobs are distributed among the different nodes through SGE queue manager (Sun Grid Engine).

Below are the available queues:

1. normal: the queue is used by default. It supports sequential and parallel jobs (using OpenMPI or MPI Platform) up to 128 cores. It supports up to 8 jobs per user and the CPU limit is 2048 days. For users asking for it, they can send jobs to 1024 cores.
2. fast: For the performance of short-term jobs, sequential and parallel jobs up to 128 cores. The limit is of 2 hours of work wallclock.
3. interactive: for the execution of interactive jobs (compilations, etc.)
4. sva: the visualization cluster. Sequential or parallel jobs to a maximun of 72 cores.
5. gpu: GPU`s cluster, sequential or parallel jobs that use CUDA GPU`s.
6. gpu_dev: system development environment for CUDA.
7. bigmem: sequential jobs OpenMP (up to 16 cores) and MPI (up to 80 cores) in systems with a memory ratio 8GB/core (128 GB)
8. hugemem: Similar than the previous, but with a ratio of 16 GB/core.
9. sla: For customers with special arrangements of service quality.

To access to the service calculation is necessary to create a group containing at least one user.

Storage

The gropus and users can request a storage space for their exclusive use.

Cloud Computing

There are two types of virtualization services:

• Pre-configured instances: The user can hire part-time access to virtual machines with predetermined features.
• Instances of the letter: The user can configure a virtual machine with the exact requirements needed.

Other services

• Hosting of systems: For those organizations, connected to the academic communication network, that they want to host their physical systems. The managed service includes managing supply, network port and a private IP address.
• VPN Connection
• Technical Support: When users require specialized services may use technical assistance Analyst Operator HPC systems and HPC.
• Special Services: The following services are available which should be personalized offer:
  • Special configurations
  • Support Centers
  • Migration of programs
  • Development and/or optimization of programs.
  • HPC consulting