(flink) branch master updated: [FLINK-39879][tests] Avoid deploy RPC timeout in CheckpointAcknowledgeFailureITCase (#28377)

[lihaosky]

This is an automated email from the ASF dual-hosted git repository.

lihaosky pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/flink.git


The following commit(s) were added to refs/heads/master by this push:
     new 5338ead2a0f [FLINK-39879][tests] Avoid deploy RPC timeout in 
CheckpointAcknowledgeFailureITCase (#28377)
5338ead2a0f is described below

commit 5338ead2a0f1f30a051e086ff3c50d165ed2b2b8
Author: Martijn Visser <2989614+martijnvis...@users.noreply.github.com>
AuthorDate: Mon Jun 15 08:49:10 2026 -0700

    [FLINK-39879][tests] Avoid deploy RPC timeout in 
CheckpointAcknowledgeFailureITCase (#28377)
---
 .../test/checkpointing/CheckpointAcknowledgeFailureITCase.java     | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git 
a/flink-tests/src/test/java/org/apache/flink/test/checkpointing/CheckpointAcknowledgeFailureITCase.java
 
b/flink-tests/src/test/java/org/apache/flink/test/checkpointing/CheckpointAcknowledgeFailureITCase.java
index 70213a1ce77..17b2f6ab23b 100644
--- 
a/flink-tests/src/test/java/org/apache/flink/test/checkpointing/CheckpointAcknowledgeFailureITCase.java
+++ 
b/flink-tests/src/test/java/org/apache/flink/test/checkpointing/CheckpointAcknowledgeFailureITCase.java
@@ -68,8 +68,11 @@ public class CheckpointAcknowledgeFailureITCase extends 
TestLogger {
     // let all the state go via checkpoint ACK RPC to exceed frame size limit
     private static final MemorySize IN_MEM_STATE_THRESHOLD = new 
MemorySize(STATE_SIZE * 2);
 
-    // let pekko time out checkpoint ACK
-    private static final Duration ASK_TIMEOUT = Duration.ofMillis(250);
+    // The oversized checkpoint ACK exceeds the pekko frame size, so it can 
never be delivered and
+    // always surfaces as an AskTimeoutException; this ask timeout only 
controls how quickly that
+    // surfaces, not whether the ACK fails. Because it is the cluster-wide RPC 
timeout it also
+    // covers task deployment, so keep it large enough that deployment does 
not time out on slow CI.
+    private static final Duration ASK_TIMEOUT = Duration.ofSeconds(1);
     // do NOT let flink time out the checkpoint
     private static final Duration CHECKPOINT_TIMEOUT = 
ASK_TIMEOUT.multipliedBy(1000);