Hi flink users, We have a problem and think flink may be a good solution for that. But I'm new to flink and hope can get some insights from flink community :)
Here is the problem. Suppose we have a DynamoDB table which store the inventory data, the schema is like: * vendorId (primary key) * inventory name * inventory units * inbound time ... This DDB table keeps changing, since we have inventory coming and removal. *Every change will trigger a DynamoDB stream. * We need to calculate *all the inventory units that > 15 days for a specific vendor* like this: > select vendorId, sum(inventory units) > from dynamodb > where today's time - inbound time > 15 > group by vendorId We don't want to schedule a daily batch job, so we are trying to work on a micro-batch solution in Flink, and publish this data to another DynamoDB table. A draft idea is to use the total units minus <15 days units, since both of then have event trigger. But no detailed solutions yet. Could anyone help provide some insights here? Thanks, J.
