We’re using the order API to download orders from a customer store. The connector polls the order API endpoint every 15 minutes and downloads the latest 50 orders. If the order already exists in our system, we skip it. That part is working fine. The issue is that the occasional order is being missed.
They have some customers that order online, and then pay via bank deposit days later (or longer) - at which time admin set the order status to “paid” manually.
Obviously our connector is missing these orders because the “created_date” is a week earlier or more - and has moved on.
We could use the “updated_at_min” operator - which should catch it when the “financial_status” changes to “paid” at which time the “updated_at” date would mean the transaction would (should?)
be caught by our scheduled download. However, it seems to me that if someone adds a comment to an older order - that would change the “update_at” date which would move it into our download “window” (when we already have this order and dont need to scrape it again).
We’re not using webhooks because they tend to be unreliable.
Can someone tell me the best practise here for catching these older orders?