ignite-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Semen Boikov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (IGNITE-1524) Client node throws exception when server node killed
Date Tue, 22 Sep 2015 14:08:04 GMT

    [ https://issues.apache.org/jira/browse/IGNITE-1524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14902689#comment-14902689
] 

Semen Boikov commented on IGNITE-1524:
--------------------------------------

Found error in client reconnect processing (ServerImpl.processClientReconnectMessage): client
reconnect is processed by coordinator node, if client was connected to coordinator and coordinator
failed but client reconnect message is handled before coordinator node failed message, then
TcpDiscoveryClientReconnectMessage is not really handled and sent back to client with 'success'
flag set to 'false'. Fixed this to not sent response to client until message is really handled
by coordinator.

> Client node throws exception when server node killed
> ----------------------------------------------------
>
>                 Key: IGNITE-1524
>                 URL: https://issues.apache.org/jira/browse/IGNITE-1524
>             Project: Ignite
>          Issue Type: Bug
>            Reporter: Artem Shutak
>            Assignee: Semen Boikov
>             Fix For: ignite-1.4
>
>         Attachments: logs-20150922-120712.zip
>
>
> In process of working on IGNITE-1397 I've found that Ignite client node throws CacheException
which caused by IgniteClientDisconnectedException (stack trace below) sometimes when I kill
one server node at cluster.
> My configuration:
> - 3 server nodes
> - 2 client nodes
> - discovery configuration:
> {code}
>         <property name="discoverySpi">
>             <bean class="org.apache.ignite.spi.discovery.tcp.TcpDiscoverySpi">
>                 <property name="ipFinder">
>                     <bean class="org.apache.ignite.spi.discovery.tcp.ipfinder.vm.TcpDiscoveryVmIpFinder">
>                         <property name="addresses">
>                             <list>
>                                 <value>127.0.0.1:47500</value>
>                                 <value>127.0.0.1:47501</value>
>                                 <value>127.0.0.1:47502</value>
>                             </list>
>                         </property>
>                     </bean>
>                 </property>
>             </bean>
>         </property>
> {code}
> Test description:
> - client nodes make random atomic ops to cluster (pts, gets, invoke, remove)
> - the first client node has a thread which kill with -9 and start one server one by one.
Kills every 60 seconds and run after 10 seconds.
> Client node fails not always. 
> {noformat}
> javax.cache.CacheException: class org.apache.ignite.IgniteClientDisconnectedException:
Operation has been cancelled (client node disconnected).
> 	at org.apache.ignite.internal.processors.cache.GridCacheUtils.convertToCacheException(GridCacheUtils.java:1621)
> 	at org.apache.ignite.internal.processors.cache.IgniteCacheProxy.cacheException(IgniteCacheProxy.java:1749)
> 	at org.apache.ignite.internal.processors.cache.IgniteCacheProxy.get(IgniteCacheProxy.java:815)
> 	at org.apache.ignite.yardstick.cache.failover.IgniteAtomicRetriesFailoverBenchmark.test(IgniteAtomicRetriesFailoverBenchmark.java:39)
> 	at org.yardstickframework.impl.BenchmarkRunner$2.run(BenchmarkRunner.java:165)
> 	at java.lang.Thread.run(Thread.java:745)
> Caused by: class org.apache.ignite.IgniteClientDisconnectedException: Operation has been
cancelled (client node disconnected).
> 	at org.apache.ignite.internal.util.IgniteUtils$14.apply(IgniteUtils.java:801)
> 	at org.apache.ignite.internal.util.IgniteUtils$14.apply(IgniteUtils.java:799)
> 	... 6 more
> Caused by: class org.apache.ignite.internal.IgniteClientDisconnectedCheckedException:
Operation has been cancelled (client node disconnected).
> 	at org.apache.ignite.internal.processors.cache.GridCacheMvccManager.disconnectedError(GridCacheMvccManager.java:360)
> 	at org.apache.ignite.internal.processors.cache.GridCacheMvccManager.onDisconnected(GridCacheMvccManager.java:334)
> 	at org.apache.ignite.internal.processors.cache.GridCacheSharedContext.onDisconnected(GridCacheSharedContext.java:141)
> 	at org.apache.ignite.internal.processors.cache.GridCacheProcessor.onDisconnected(GridCacheProcessor.java:997)
> 	at org.apache.ignite.internal.IgniteKernal.onDisconnected(IgniteKernal.java:2983)
> 	at org.apache.ignite.internal.managers.discovery.GridDiscoveryManager$5.onDiscovery(GridDiscoveryManager.java:592)
> 	at org.apache.ignite.spi.discovery.tcp.ClientImpl$MessageWorker.notifyDiscovery(ClientImpl.java:2024)
> 	at org.apache.ignite.spi.discovery.tcp.ClientImpl$MessageWorker.notifyDiscovery(ClientImpl.java:2005)
> 	at org.apache.ignite.spi.discovery.tcp.ClientImpl$MessageWorker.body(ClientImpl.java:1422)
> 	at org.apache.ignite.spi.IgniteSpiThread.run(IgniteSpiThread.java:62)
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message