Maintenance: empty-trash as a LRT

[pcrespov]

          I will comment on it here. Your current will fail at some point, due to random conditions or load, this what I got form from your description on the PR.

If I also understood the baseline issue right, the frontend would be blocked for too much time and we'd like to avoid this.

Why not use what it's already been proven to work. Long running jobs. We have support for them currently. The frontend will know when this ends and it does not have to block. It can poll/wait for the result. Would it not work for this case?

Originally posted by @GitHK in #7416 (comment)

NOTE: this is an improvement on #7379

[pcrespov]

They are called async jobs and are currently only added to storage, the interface should be extracted and shared with other services as well. That's the current issue

• https://github.com/ITISFoundation/osparc-simcore/blob/master/services/storage/src/simcore_service_storage/api/rpc/_async_jobs.py

the only ones that are usable as is are the ones via http exposed in servicelib