Considering having a meta-command to pull output directory of a job then initiates a deletion process #37
Description
I'm only just realizing that to delete a job output folder in Digital Ocean, it moves folders into "pending destruction state," which then takes 2 calendar weeks before it actually deletes. This means that even through the DigitalOcean API, jobs are stuck in "deleting" mode
Thus, I think it would help to have a command that would effectively run:
cloudexec pull --job XXX output # pulls output directory
cloudexec clean --job XXX # removes all files from that job
It would mean every time a job finishes, we can immediately pull the output, and start a process deletion which doesn't allow the state to grow so crazily.