Github user dawidwys commented on a diff in the pull request:
https://github.com/apache/flink/pull/6283#discussion_r202098798
--- Diff:
flink-runtime/src/test/java/org/apache/flink/runtime/checkpoint/CoordinatorShutdownTest.java
---
@@ -83,6 +86,10 @@ public void testCoordinatorShutsDownOnFailure() {
CheckpointRetentionPolicy.NEVER_RETAIN_AFTER_TERMINATION,
true),
null));
+
+ ExecutionConfig executionConfig = new ExecutionConfig();
+
executionConfig.setRestartStrategy(RestartStrategies.fallBackRestart());
--- End diff --
Right now null is a bit different than `FallbackRestartStrategy`.
* null - allows fallback to `FixedRestartStrategy` in case of checkpointing
enabled and `noRestart` was set on server-side
* `FallbackRestartStrategy` - always the server-side strategy is used
(indifferent to checkpointing)
If we by default set the `FallbackStrategy` we have two options:
* we either always set `FixedRestartStrategy` if checkpointing is enabled
and `noRestart` was set on server side
* we never automatically fallback to `FixedRestartStrategy`, even in case
of checkpointing.
What do you think would be better option? Keep the null, always fallback to
`FixedRestartStrategy` or never fallback to it?
---
