You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Brandon Williams (Jira)" <ji...@apache.org> on 2021/02/18 22:54:00 UTC

[jira] [Comment Edited] (CASSANDRA-16358) Minor Flakiness in ProxyHandlerConnectionsTest#testExpireSomeFromBatch

    [ https://issues.apache.org/jira/browse/CASSANDRA-16358?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17286758#comment-17286758 ] 

Brandon Williams edited comment on CASSANDRA-16358 at 2/18/21, 10:53 PM:
-------------------------------------------------------------------------

There are a couple problems with this test.  The first that if messages are dropped, which can often be expected, it results in an NPE that complicates further debugging.  The timeout for this test itself does not actually matter, since it is manipulating timing windows that it later assert results upon.  These all work in a perfect world, but inside the JVM need some relaxing to allow for the real world to get in the way. Patch [here|https://github.com/driftx/cassandra/tree/CASSANDRA-16358], [CI here|https://ci-cassandra.apache.org/job/Cassandra-devbranch/388/], that addresses both of these.  After being able to reproduce the issue within hundreds of runs, I've had success in thousands with this.


was (Author: brandon.williams):
There are a couple problems with this test.  The first that if messages are dropped, which can often be expected, it results in an NPE that complicates further debugging.  The timeout for this test itself does not actually matter, since it is manipulating timing windows itself that it later assert results upon.  These all work in a perfect world, but inside the JVM need some relaxing to allow for the real world to get in the way. Patch [here|https://github.com/driftx/cassandra/tree/CASSANDRA-16358], [CI here|https://ci-cassandra.apache.org/job/Cassandra-devbranch/388/], that addresses both of these.  After being able to reproduce the issue within hundreds of runs, I've had success in thousands with this.

> Minor Flakiness in ProxyHandlerConnectionsTest#testExpireSomeFromBatch
> ----------------------------------------------------------------------
>
>                 Key: CASSANDRA-16358
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-16358
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Test/unit
>            Reporter: Caleb Rackliffe
>            Assignee: Brandon Williams
>            Priority: Normal
>             Fix For: 4.0, 4.0-beta
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> This seems only to be failing in ci-cassandra, but it fails across cdc, compression, and normal configurations there. The only failure recent enough to be retained is here:
> https://ci-cassandra.apache.org/job/Cassandra-devbranch/268/testReport/org.apache.cassandra.net/ProxyHandlerConnectionsTest/testExpireSomeFromBatch_compression_2/
> {noformat}
> org.apache.cassandra.net.ProxyHandlerConnectionsTest.testExpireSomeFromBatch-compression (from org.apache.cassandra.net.ProxyHandlerConnectionsTest-compression)
> Error Message
> Expired: 8, Arrived: 10
> Stacktrace
> junit.framework.AssertionFailedError: Expired: 8, Arrived: 10
> 	at org.apache.cassandra.net.ProxyHandlerConnectionsTest.waitForCondition(ProxyHandlerConnectionsTest.java:285)
> 	at org.apache.cassandra.net.ProxyHandlerConnectionsTest.lambda$testExpireSomeFromBatch$23(ProxyHandlerConnectionsTest.java:210)
> 	at org.apache.cassandra.net.ProxyHandlerConnectionsTest.doTestManual(ProxyHandlerConnectionsTest.java:378)
> 	at org.apache.cassandra.net.ProxyHandlerConnectionsTest.testManual(ProxyHandlerConnectionsTest.java:337)
> 	at org.apache.cassandra.net.ProxyHandlerConnectionsTest.testExpireSomeFromBatch(ProxyHandlerConnectionsTest.java:169)
> 	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> 	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> 	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> Standard Output
> DEBUG [main] 2020-12-15 20:36:43,018 InternalLoggerFactory.java:45 - Using SLF4J as the default logging framework
> DEBUG [main] 2020-12-15 20:36:43,043 PlatformDependent0.java:417 - -Dio.netty.noUnsafe: false
> DEBUG [main] 2020-12-15 20:36:43,044 PlatformDependent0.java:897 - Java version: 8
> DEBUG [main] 2020-12-15 20:36:43,044 PlatformDependent0.java:130 - sun.misc.Unsafe.theUnsafe: available
> DEBUG [main] 2020-12-15 20:36:43,045 PlatformDependent0.java:154 - sun.misc.Unsafe.copyMemory: available
> ...[truncated 394654 chars]...
> ol$Initiate.maybeDecode(HandshakeProtocol.java:167)
> 	at org.apache.cassandra.net.InboundConnectionInitiator$Handler.initiate(InboundConnectionInitiator.java:242)
> 	at org.apache.cassandra.net.InboundConnectionInitiator$Handler.decode(InboundConnectionInitiator.java:235)
> 	at io.netty.handler.codec.ByteToMessageDecoder.decodeRemovalReentryProtection(ByteToMessageDecoder.java:501)
> 	at io.netty.handler.codec.ByteToMessageDecoder.callDecode(ByteToMessageDecoder.java:440)
> 	... 29 common frames omitted
> {noformat}
> There are a number of historical failures listed here, but the runs that produced them have not been retained:
> https://ci-cassandra.apache.org/job/Cassandra-trunk/207/testReport/junit/org.apache.cassandra.net/ProxyHandlerConnectionsTest/



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org