[jira] [Commented] (GEODE-8131) alert service hangs, blocking cache operations

["Kirk Lund (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/GEODE-8131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17124318#comment-17124318 ] 

Kirk Lund commented on GEODE-8131:
----------------------------------

Assigned to Bruce to provide me with more info.

I need some more info on what’s wrong with Alerting and how to fix it. The main changes I made to Alerting recently were:
1. Move the implementation that uses log4j-core to geode-log4j
2. Use a dedicated ExecutorService in the sender to send an AlertListenerMessage

The receiving side of Alerts didn’t change other than the reorg to geode-log4j. So I’m assuming GEODE-8131 is an existing bug that has always existed in Geode.

If you want me to start using an Executor on the receiving side, we can do that. If you want the Alert message to be a different type, we can do that too.

I vaguely remember something about different message types coming into different Executors in DM. But, it’s been years since I’ve worked on messages like this so I need some detailed info on what’s wrong and how to fix it. Thanks!

> alert service hangs, blocking cache operations
> ----------------------------------------------
>
>                 Key: GEODE-8131
>                 URL: https://issues.apache.org/jira/browse/GEODE-8131
>             Project: Geode
>          Issue Type: Bug
>          Components: logging
>            Reporter: Bruce J Schuchardt
>            Assignee: Bruce J Schuchardt
>            Priority: Major
>              Labels: GeodeOperationAPI
>
> This v1.8 TcpConduit reader thread was blocked in a production system.  It had experienced a deserialization error and was trying to log the exception.  A Manager was present in the cluster and had registered as an alert listener.  Another thread was blocked sending something on the shared/unordered connection that this alert should be sent on.  This persisted for over 6 hours and we never saw the serialization exception in the log file.  Consequently we had to recommend setting the alert level to None and have them run into the serialization problem again.
> This is a serious flaw in the alerting system and it's caused us grief many times.  The alerting system should not block other threads.  Maybe a background thread could consume and transmit alerts to alert-listeners?
>  
> {noformat}
> "P2P message reader for 10.236.28.120(servername-removed)<v491>:56152 shared unordered uid=9 port=41204" tid=0xd49 (in native)    java.lang.Thread.State: RUNNABLE at sun.nio.ch.FileDispatcherImpl.write0(Native Method) at sun.nio.ch.SocketDispatcher.write(SocketDispatcher.java:47) at sun.nio.ch.IOUtil.writeFromNativeBuffer(IOUtil.java:93) at sun.nio.ch.IOUtil.write(IOUtil.java:51) at sun.nio.ch.SocketChannelImpl.write(SocketChannelImpl.java:471) -  locked java.lang.Object@24528b9b at org.apache.geode.internal.tcp.Connection.nioWriteFully(Connection.java:3291) -  locked java.lang.Object@42a1a79b at org.apache.geode.internal.tcp.Connection.sendPreserialized(Connection.java:2527) at org.apache.geode.internal.tcp.MsgStreamer.realFlush(MsgStreamer.java:319) at org.apache.geode.internal.tcp.MsgStreamer.writeMessage(MsgStreamer.java:244) at org.apache.geode.distributed.internal.direct.DirectChannel.sendToMany(DirectChannel.java:393) at org.apache.geode.distributed.internal.direct.DirectChannel.sendToOne(DirectChannel.java:250) at org.apache.geode.distributed.internal.direct.DirectChannel.send(DirectChannel.java:615) at org.apache.geode.distributed.internal.membership.gms.mgr.GMSMembershipManager.directChannelSend(GMSMembershipManager.java:1717) at org.apache.geode.distributed.internal.membership.gms.mgr.GMSMembershipManager.send(GMSMembershipManager.java:1898) at org.apache.geode.distributed.internal.ClusterDistributionManager.sendViaMembershipManager(ClusterDistributionManager.java:2878) at org.apache.geode.distributed.internal.ClusterDistributionManager.sendOutgoing(ClusterDistributionManager.java:2798) at org.apache.geode.distributed.internal.ClusterDistributionManager.sendMessage(ClusterDistributionManager.java:2837) at org.apache.geode.distributed.internal.ClusterDistributionManager.putOutgoing(ClusterDistributionManager.java:1531) at org.apache.geode.internal.alerting.AlertMessaging.sendAlert(AlertMessaging.java:75) at org.apache.geode.internal.logging.log4j.AlertAppender.sendAlertMessage(AlertAppender.java:188) at org.apache.geode.internal.logging.log4j.AlertAppender.doAppend(AlertAppender.java:163) at org.apache.geode.internal.logging.log4j.AlertAppender.lambda$append$0(AlertAppender.java:159) at org.apache.geode.internal.logging.log4j.AlertAppender$$Lambda$168/1102181662.run(Unknown Source) at org.apache.geode.internal.alerting.AlertingAction.execute(AlertingAction.java:29) at org.apache.geode.internal.logging.log4j.AlertAppender.append(AlertAppender.java:159) at org.apache.logging.log4j.core.config.AppenderControl.tryCallAppender(AppenderControl.java:156) at org.apache.logging.log4j.core.config.AppenderControl.callAppender0(AppenderControl.java:129) at org.apache.logging.log4j.core.config.AppenderControl.callAppenderPreventRecursion(AppenderControl.java:120) at org.apache.logging.log4j.core.config.AppenderControl.callAppender(AppenderControl.java:84) at org.apache.logging.log4j.core.config.LoggerConfig.callAppenders(LoggerConfig.java:464) at org.apache.logging.log4j.core.config.LoggerConfig.processLogEvent(LoggerConfig.java:448) at org.apache.logging.log4j.core.config.LoggerConfig.log(LoggerConfig.java:431) at org.apache.logging.log4j.core.config.LoggerConfig.logParent(LoggerConfig.java:455) at org.apache.logging.log4j.core.config.LoggerConfig.processLogEvent(LoggerConfig.java:450) at org.apache.logging.log4j.core.config.LoggerConfig.log(LoggerConfig.java:431) at org.apache.logging.log4j.core.config.LoggerConfig.log(LoggerConfig.java:406) at org.apache.logging.log4j.core.config.AwaitCompletionReliabilityStrategy.log(AwaitCompletionReliabilityStrategy.java:63) at org.apache.logging.log4j.core.Logger.logMessage(Logger.java:146) at org.apache.logging.log4j.spi.ExtendedLoggerWrapper.logMessage(ExtendedLoggerWrapper.java:217) at org.apache.logging.log4j.spi.AbstractLogger.tryLogMessage(AbstractLogger.java:2170) at org.apache.logging.log4j.spi.AbstractLogger.logMessageTrackRecursion(AbstractLogger.java:2125) at org.apache.logging.log4j.spi.AbstractLogger.logMessageSafely(AbstractLogger.java:2108) at org.apache.logging.log4j.spi.AbstractLogger.logMessage(AbstractLogger.java:2002) at org.apache.logging.log4j.spi.AbstractLogger.logIfEnabled(AbstractLogger.java:1974) at org.apache.logging.log4j.spi.AbstractLogger.fatal(AbstractLogger.java:1054) at org.apache.geode.internal.tcp.Connection.processNIOBuffer(Connection.java:3610) at org.apache.geode.internal.tcp.Connection.runNioReader(Connection.java:1824) at org.apache.geode.internal.tcp.Connection.run(Connection.java:1686) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) {noformat}



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