The access-request-status API is falling short for meeting needs as follows.
When a significant outage occurred, sources failed in such a way that there are many requests in flight which will never finish. Unfortunately, that means managers are seeing lost requests.
I’ve been tasked with identifying these requests and cancelling them via the API.
There are a few ways I might retrieve a list of requests and begin finding the offending ones.
- Filter for requests since the date of the outage
- Find requests which include a specific source
- Find requests in a certain status
- Load every single request that has ever been made in our system and filter manually from this list of tens of thousands of requests.
The API documentation also shows there are some query fields available along with filter fields (which seems to only allow one property to be filtered). Why does this API use separate queries and filters? There are sorters also, which aren’t part of the filters.