I am currently using the Extract API for an integration . After issuing an extract job request I loop the status link until the file is ready to collect. In our implementation environment this was quite quick however in production it is exceeding my codes hard timeout of 3 hours. most of the time the file that is produced only contains the header. I assume this is because my request is placed on a shared queue across multiple tenants. Rather than locking the status check indefinitely is there a recommended pattern or length of time for checking the job status?