[GitHub] [flink] pnowojski commented on a diff in pull request #19907: [FLINK-27692][state] Support local recovery for materialized part of changelog

[GitBox <gi...@apache.org>]

pnowojski commented on code in PR #19907:
URL: https://github.com/apache/flink/pull/19907#discussion_r903436217


##########
flink-runtime/src/main/java/org/apache/flink/runtime/taskexecutor/TaskExecutor.java:
##########
@@ -685,13 +687,27 @@ public CompletableFuture<Acknowledge> submitTask(
                     jobManagerConnection.getClassLoaderHandle();
             PartitionProducerStateChecker partitionStateChecker =
                     jobManagerConnection.getPartitionStateChecker();
+            TernaryBoolean changelogEnabled = TernaryBoolean.FALSE;
+            try {
+                changelogEnabled =
+                        InstantiationUtil.readObjectFromConfig(
+                                taskInformation.getTaskConfiguration(),
+                                "enablechangelog",
+                                getClass().getClassLoader());

Review Comment:
   I'm not sure, but I think the idea sometime ago was to make "streaming api" runtime free. There were also independent plans of merging streaming runtime and batch runtime (or basically getting rid of batch runtime with dropping `DataSet API`). All in all it seems to me like `flink-core` would be a better place for config options? Indeed it would be nice to hear what @zentol thinks about it, as I presume/hope he has a better knowledge about those issues.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscribe@flink.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org