Roman Puchkovskiy created IGNITE-18292:
------------------------------------------
Summary: Ignite 3 cluster sometimes hangs making
KeyValueViewPocoTests.TestContains fail
Key: IGNITE-18292
URL: https://issues.apache.org/jira/browse/IGNITE-18292
Project: Ignite
Issue Type: Bug
Reporter: Roman Puchkovskiy
Assignee: Roman Puchkovskiy
Fix For: 3.0.0-beta2
The corresponding test fails regularly:
[https://ci.ignite.apache.org/test/-8907008159462326767?currentProjectId=ApacheIgnite3xGradle_Test&expandTestHistoryChartSection=true&orderBy=status&order=desc]
The following can be found in the log of build #6620
[https://ci.ignite.apache.org/buildConfiguration/ApacheIgnite3xGradle_Test_RunNetTests/6935941?hideProblemsFromDependencies=false&hideTestsFromDependencies=false&expandBuildProblemsSection=true&expandBuildChangesSection=true&expandBuildTestsSection=true&showLog=6935941_11284_971.988&logFilter=debug&logView=flowAware]
[16:47:18] : [dotnet test] java.lang.NullPointerException
[16:47:18] : [dotnet test] at
org.apache.ignite.network.DefaultMessagingService.send0(DefaultMessagingService.java:146)
[16:47:18] : [dotnet test] at
org.apache.ignite.network.DefaultMessagingService.respond(DefaultMessagingService.java:124)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.impl.IgniteRpcServer$RpcMessageHandler$1.sendResponse(IgniteRpcServer.java:183)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.impl.core.AppendEntriesRequestProcessor.processRequest0(AppendEntriesRequestProcessor.java:430)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.impl.core.AppendEntriesRequestProcessor.processRequest0(AppendEntriesRequestProcessor.java:41)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.impl.core.NodeRequestProcessor.processRequest(NodeRequestProcessor.java:55)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.RpcRequestProcessor.handleRequest(RpcRequestProcessor.java:49)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.RpcRequestProcessor.handleRequest(RpcRequestProcessor.java:29)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.rpc.impl.IgniteRpcServer$RpcMessageHandler.lambda$onReceived$0(IgniteRpcServer.java:195)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.util.concurrent.MpscSingleThreadExecutor$Worker.runTask(MpscSingleThreadExecutor.java:354)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.util.concurrent.MpscSingleThreadExecutor$Worker.run(MpscSingleThreadExecutor.java:338)
[16:47:18] : [dotnet test] at
org.apache.ignite.raft.jraft.util.concurrent.MpscSingleThreadExecutor.lambda$doStartWorker$3(MpscSingleThreadExecutor.java:262)
[16:47:18] : [dotnet test] at
java.base/java.lang.Thread.run(Thread.java:834)
Another interesting part of log is this:
[16:47:09] : [dotnet test] INFO:
[default:org.apache.ignite.internal.runner.app.PlatformTestNodeRunner:[email protected]:3344]
Member leaved without notification:
default:org.apache.ignite.internal.runner.app.PlatformTestNodeRunner_2:[email protected]:3345
and then
[16:47:09] : [dotnet test] INFO:
[default:org.apache.ignite.internal.runner.app.PlatformTestNodeRunner_2:[email protected]:3345]
Member leaved without notification:
default:org.apache.ignite.internal.runner.app.PlatformTestNodeRunner:[email protected]:3344
Looks like the 2 nodes comprising the cluster lost each other and were never
able to see each other again, even though they were running on the same machine
(just on different ports). I wonder could this be some network-related problem.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
