Home

GATE-Lab

How it works

Basic simulation workflow for a GATE run:

  1. input files are parsed from the local macfile, zipped and copied from local machine to a grid storage element.
  2. the simulation (i.e. input archive location, gate and merge release locations, name of the main mac file, estimated duration and total number of particles) is sent to the GATE-Lab server (host ui.egee in Creatis server room at Leonard de Vinci building).
  3. GATE-Lab server creates the tasks, start master and submit agents to grid resources if necessary. Tasks are monitored and resubmitted (up to a max number of times) when they fail.
  4. Tasks run on remote nodes, download inputs from grid storage, compute results and store them on grid storage elements.
  5. Results are browsed and copied from the grid to local machine using VBrowser.

You can download here a short demo (.exe file) of how GATE-Lab works. For the Flash version (.wnk file) click here. A short presentation of the GateLab is also available here

Warnings

  • The GateLab has a configuration tab with several variables that need to be set up appropriately. They have a default value which may not correspond to your needs. Please contact egee-admin@creatis.insa-lyon.fr for information about these variables.
  • The GateLab is still in a test phase.