You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Alexey Platonov (JIRA)" <ji...@apache.org> on 2018/09/11 08:10:00 UTC

[jira] [Commented] (IGNITE-8650) ZookeeperDiscovery testClientReconnectMultinode is flaky in master

    [ https://issues.apache.org/jira/browse/IGNITE-8650?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16610248#comment-16610248 ] 

Alexey Platonov commented on IGNITE-8650:
-----------------------------------------

CheckClientsStatus may fail because of in rare cases while reconnecting already disconnected client gets list of alive nodes without this client. Such case can be skipped on client.

> ZookeeperDiscovery testClientReconnectMultinode is flaky in master
> ------------------------------------------------------------------
>
>                 Key: IGNITE-8650
>                 URL: https://issues.apache.org/jira/browse/IGNITE-8650
>             Project: Ignite
>          Issue Type: Bug
>            Reporter: Alexey Goncharuk
>            Assignee: Alexey Platonov
>            Priority: Major
>              Labels: MakeTeamcityGreenAgain
>             Fix For: 2.8
>
>
> The test IgniteClientReconnectCacheTest#testReconnectMultinode(LongHistory) periodically fails with timeouts in master.
> From the logs I see that the hang is caused by one of the two assertion errors:
> {code}
> java.lang.AssertionError
> 	at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.checkClientsStatus(ZookeeperDiscoveryImpl.java:1345)
> 	at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.access$2300(ZookeeperDiscoveryImpl.java:108)
> 	at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl$CheckClientsStatusCallback.processResult0(ZookeeperDiscoveryImpl.java:4332)
> 	at org.apache.ignite.spi.discovery.zk.internal.ZkAbstractChildrenCallback.processResult(ZkAbstractChildrenCallback.java:42)
> 	at org.apache.ignite.spi.discovery.zk.internal.ZookeeperClient$ChildrenCallbackWrapper.processResult(ZookeeperClient.java:1132)
> 	at org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.java:590)
> 	at org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:498)
> {code}
> or 
> {code}
> java.lang.AssertionError
>     at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.checkClientsStatus(ZookeeperDiscoveryImpl.java:1388)
>     at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl.access$2300(ZookeeperDiscoveryImpl.java:108)
>     at org.apache.ignite.spi.discovery.zk.internal.ZookeeperDiscoveryImpl$CheckClientsStatusCallback.processResult0(ZookeeperDiscoveryImpl.java:4332)
>     at org.apache.ignite.spi.discovery.zk.internal.ZkAbstractChildrenCallback.processResult(ZkAbstractChildrenCallback.java:42)
>     at org.apache.ignite.spi.discovery.zk.internal.ZookeeperClient$ChildrenCallbackWrapper.processResult(ZookeeperClient.java:1132)
>     at org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.java:590)
>     at org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:498)
> {code}
> The test failure can be rarely reproduced locally (run repeatedly with CPU stress enabled).



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)