Read one bulk job

Authenticate with an operator token or an app API key or a signed-in Foliot Console session.

Poll this for progress: done and failed against total.

A job survives the restart of the server running it. If that server stops, another replica takes the job over within about a minute and continues from the last run it finished, so done keeps rising and no run is acted on twice (a run in flight at the instant of a crash can be, once). While it is handed over a job can read pending again. A job that was already stranded before this was possible cannot be placed and is marked failed with an error saying how far it got (up to one page of runs beyond that count may also have been acted on); submit it again if runs remain. A job nobody has touched for more than a day is likewise marked failed and not resumed, because a resumed job acts on the runs that match its filter at that moment, which may no longer be the ones it was submitted for.

Path parameters

Responses