Use case: pipeline multiple execution

[denisri]

Author: Yann C.

Summary: Run several time the same pipeline on the same data

Description:

This pipeline assess the ability of the system to handle several executions of the same pipeline on the same data. The main difficulties addressed would be the automatic naming of the output file names (several files generated with exactly the same metadata in inputs) and ability to identify output data according to pipeline parameter set.

The targeted pipeline could be Morphologist. It would be run three times with three different sets of parameters (TDB which parameters; for instance different normalizations). An interesting feature would be to have the possibility to add custom metadata on output data on each execution. This way the output could be sorted according to, for example, the value of my_dataset attribute that would be "first", "second" or "third".

Participants: single user

Datasets: The input dataset is not very important here but a large dataset could be interesting to assess the speed of the metadata query on output files using custom attributes.

Data flow: NA

Security: NA

Workgroup(s): pipeline, metadata