For each step in a workflow executed by the ADES, we want to capture CPU, storage, and memory metrics. This might be added to over time, and could include things like run-time, network in/out etc.
The goal of these metrics is to tune algorithms over time.
Things to consider- is this a 'flag' in the WPS execution request to enable this? or is it always running
what is the output, format, and where does it go. We could start simple and create a file for each run, or be advanced and add this info to the jobs database, provenance, or some common logging system.
For each step in a workflow executed by the ADES, we want to capture CPU, storage, and memory metrics. This might be added to over time, and could include things like run-time, network in/out etc.
The goal of these metrics is to tune algorithms over time.