Cassandra supports upsert for UPDATE
. That is, if the key provided in the update statement doesn't exist in the table; it is implied as insert. Is there a way to know from the return code of the update query if update resulted in update or insert. We would prefer to detect this without an additional read from Cassandra.
We have scenario where we have a very high throughput application where 99% of the events results in insert but a very small chunk results in update. Once updated we would want perform some additional checks and trigger some.