Sounds like your describing a source function that subscribes to couch
db updates. You'd usually implement this as a Co(Flat)MapFunction that
has 2 inputs, 1 from kafka and one from couch db, which stores the
processing parameters in state.
There's no built-in way to subscribe to couch db updates, so you'll have
to find a client to do so yourself.
On 12/04/2019 12:41, Soheil Pourbafrani wrote:
> Hi,
>
> In my problem I should Process Kafka messages Using Apache Flink,
> while some processing parameters should be read from the CouchDB, So I
> have two questions:
>
> 1- What is Flink way to read data from the CouchDB?
> 2- I want to trigger Flink to load data from the Couch DB if a new
> document was inserted into the table, instead of reading data from the
> database in every message processing. What should be my strategy in
> such an environment? I think I need some trigger or event handler?
> Does Flink provide any facilities for this?
>
> Thanks