zhuzhurk commented on a change in pull request #9778: [FLINK-14206][runtime]
Let fullRestart metric count both full restarts and fine grained restarts
URL: https://github.com/apache/flink/pull/9778#discussion_r336335410
##########
File path:
flink-runtime/src/test/java/org/apache/flink/runtime/executiongraph/ExecutionGraphRestartTest.java
##########
@@ -595,9 +594,6 @@ public void
slotPoolExecutionGraph_ConcurrentSchedulingAndAllocationFailure_Shou
@Test
public void testRestartWithEagerSchedulingAndSlotSharing() throws
Exception {
- // this test is inconclusive if not used with a proper
multi-threaded executor
Review comment:
The executor(futureExecutor) is not used for scheduling/failover anymore
since '[FLINK-11417] Make access to ExecutionGraph single threaded from
JobMaster main thread' is merged.
I think there is no need to specify the executor now and the tests does not
rely on that executor to do any thing.
The idea of testRestartWithEagerSchedulingAndSlotSharing is easy that it
tests a eager scheduling job can recover from a task failure.
The idea of testRestartWithSlotSharingAndNotEnoughResources is also easy
that the job will keep failing due to not enough resource until the restart
strategy suppresses the restarts.
So I think we can just remove the assumption and the executor specification.
Like in 6d323bf2657f8f82eac28d2f529fa1dbe36fcde6.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
