Retrieve fallback schedules

[Flix6x]

FlexMeasures 0.17 will introduce a new way to handle fallback schedules. A separate fallback job will be triggered upon failure of the main scheduling job (whereas before the fallback schedule was computed within the same job, which is a less atomic transaction that complicated debugging).

When the FlexMeasures host upgrades to run flexmeasures>=0.17 the API will return a status code 303 when requesting an originally failed schedule, with the location of the fallback schedule as a response header. This behavior requires a backwards compatible update in the flexmeasures_client.py, by handling the 303 status code and implementing a follow-up action to retrieve the fallback schedule . Relevant documentation will be made available, see FlexMeasures/flexmeasures#858.