R&D Investigate on pilot lifetime projected onto HL-LHC era

[khurtado]

Impact of the new feature Place holder for R&D issue

Is your feature request related to a problem? Please describe. This is related to the list of TODO tasks in the Evaluation of the WM system for the HL-LHC scenario google document

Describe the solution you'd like The current pilot lifetime is set to 8 hours. We should investigate if changing this would be benefitial during HL-LHC era, either globally or for e.g.: HPC resources for example (e.g.: 48 hour pilots). If 8h needs to be changed, we should describe what needs to change; for instance this can be done in the reqmgr2 (but it is static now, though making it configurable would be trivial)

Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered.

Additional context Add any other context or screenshots about the feature request here.

[amaltaro]

This is actually a great idea, Kenyi. I would suggest an evaluation of a few workflows targetting job wallclocktime of: 4h, 8h, 12h, 16h.

A few metrics that come to my mind would be:

• avg job wallclocktime
• min job wallclocktime
• max job wallclocktime
• total workflow wallclocktime
• workflow turnaround (time from acquired to completed status)
• condor retry/failure rate
• wmagent retry/failure rate
• ?

@khurtado Kenyi, can you please apply the relevant labels (I guess only R&D) and fields of the project board (for the other R&D as well).