You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by "fml2 (Jira)" <ji...@apache.org> on 2020/11/13 22:27:00 UTC

[jira] [Created] (KAFKA-10722) Timestamped store is used even if not desired

fml2 created KAFKA-10722:
----------------------------

             Summary: Timestamped store is used even if not desired
                 Key: KAFKA-10722
                 URL: https://issues.apache.org/jira/browse/KAFKA-10722
             Project: Kafka
          Issue Type: Bug
          Components: streams
    Affects Versions: 2.6.0, 2.4.1
            Reporter: fml2


I have a stream which I then group and aggregate (this results in a KTable). When aggregating, I explicitly tell to materialize the result table using a usual (not timestamped) store.

After that, the KTable is filtered and streamed. This stream is processed by a processor that accesses the store.

The problem/bug ist that even if I tell to use a non-timestamped store, a timestamoed one is used, which leads to a ClassCastException.

Here is a schema of the code:
{code:java}
KTable table = ...aggregate(
  initializer, // initializer for the KTable row
  aggregator, // aggregator
  Materialized.as(Stores.persistentKeyValueStore("MyStore")) // <-- Non-Timestamped!
    .withKeySerde(...).withValueSerde(...));
{code}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)