[RATIS-2105] TestRetryCacheWithNettyRpc may fail with TimeoutException

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Assignee: Unassigned
• Reporter: Tsz-wo Sze
• Created: 02/Jun/24 23:48
• Updated: 17/Jul/24 16:22

Description

I can reproduce it with ratis-3.1.0-rc0

[INFO] Running org.apache.ratis.netty.TestRetryCacheWithNettyRpc
[ERROR] Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 100.299 s <<< FAILURE! - in org.apache.ratis.netty.TestRetryCacheWithNettyRpc
[ERROR] org.apache.ratis.netty.TestRetryCacheWithNettyRpc.testRetryOnNewLeader  Time elapsed: 100.182 s  <<< ERROR!
java.util.concurrent.TimeoutException: testRetryOnNewLeader() timed out after 100 seconds
	at java.util.ArrayList.forEach(ArrayList.java:1259)
	at java.util.ArrayList.forEach(ArrayList.java:1259)
	Suppressed: java.io.InterruptedIOException: retry policy=RetryForeverWithSleep(sleepTime = 100ms)
		at org.apache.ratis.client.impl.BlockingImpl.sendRequestWithRetry(BlockingImpl.java:138)
		at org.apache.ratis.client.impl.AdminImpl.setConfiguration(AdminImpl.java:46)
		at org.apache.ratis.client.api.AdminApi.setConfiguration(AdminApi.java:51)
		at org.apache.ratis.client.api.AdminApi.setConfiguration(AdminApi.java:45)
		at org.apache.ratis.server.impl.MiniRaftCluster.setConfiguration(MiniRaftCluster.java:826)
		at org.apache.ratis.RetryCacheTests.lambda$runTestRetryOnNewLeader$0(RetryCacheTests.java:145)
		at org.apache.ratis.server.impl.RaftServerTestUtil.runWithMinorityPeers(RaftServerTestUtil.java:231)
		at org.apache.ratis.RetryCacheTests.runTestRetryOnNewLeader(RetryCacheTests.java:144)
		at org.apache.ratis.server.impl.MiniRaftCluster$Factory$Get.runWithNewCluster(MiniRaftCluster.java:143)
		at org.apache.ratis.server.impl.MiniRaftCluster$Factory$Get.runWithNewCluster(MiniRaftCluster.java:121)
		at org.apache.ratis.RetryCacheTests.testRetryOnNewLeader(RetryCacheTests.java:124)
		at java.lang.reflect.Method.invoke(Method.java:498)
		... 2 more

Issue Links

• relates to: RATIS-2127 TestRetryCacheWithGrpc may fail with object already completely released.