[jboss-jira] [JBoss JIRA] (WFLY-7418) Batch deployments with a large number of executed jobs can lock up or slow down the web console

[Dinesh Mahadevan (Jira)]

    [ https://issues.jboss.org/browse/WFLY-7418?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13807804#comment-13807804 ] 

Dinesh Mahadevan commented on WFLY-7418:
----------------------------------------

Hi [~timonz] 
+1 on the issue.
Using WildFly 16, we have an application that fails to start until we added -Djboss.as.management.blocking.timeout=6000" allowing the application to start after 2 hours. 
Line 581 of class org.jberet.repository.JdbcRepository is the line that appears in almost all thread dumps. Judging from the context, it is executing the SQL from sqls.getProperty(SELECT_JOB_EXECUTIONS_BY_JOB_INSTANCE_ID)

The quickest solution we can find so far is to disable the MicroProfile setting on Wildfly, or purge the job_execution table manually.

{code:java}
<extension module="org.wildfly.extension.elytron"/>
<extension module="org.wildfly.extension.io"/>
<!--<extension module="org.wildfly.extension.microprofile.config-smallrye"/>
<extension module="org.wildfly.extension.microprofile.health-smallrye"/>
<extension module="org.wildfly.extension.microprofile.metrics-smallrye"/>
<extension module="org.wildfly.extension.microprofile.opentracing-smallrye"/>-->
<extension module="org.wildfly.extension.request-controller"/>
<extension module="org.wildfly.extension.security.manager"/>

 <subsystem xmlns="urn:jboss:domain:mail:3.0">
    <mail-session jndi-name="java:jboss/mail/Default" name="default">
        <smtp-server outbound-socket-binding-ref="mail-smtp"/>
    </mail-session>
</subsystem>
<!--<subsystem xmlns="urn:wildfly:microprofile-config-smallrye:1.0"/>
<subsystem xmlns="urn:wildfly:microprofile-health-smallrye:1.0" security-enabled="false"/>
<subsystem xmlns="urn:wildfly:microprofile-metrics-smallrye:2.0" exposed-subsystems="*" prefix="${wildfly.metrics.prefix:wildfly}" security-enabled="false"/>
<subsystem xmlns="urn:wildfly:microprofile-opentracing-smallrye:1.0"/>-->
<subsystem xmlns="urn:jboss:domain:naming:2.0">
    <remote-naming/>
</subsystem>
{code}


> Batch deployments with a large number of executed jobs can lock up or slow down the web console
> -----------------------------------------------------------------------------------------------
>
>                 Key: WFLY-7418
>                 URL: https://issues.jboss.org/browse/WFLY-7418
>             Project: WildFly
>          Issue Type: Enhancement
>          Components: Batch, Web Console
>            Reporter: James Perkins
>            Assignee: James Perkins
>            Priority: Major
>              Labels: move_to_halnext
>
> Batch deployments which contain a large number of executed jobs can be extremely slow to process as the {{/deployment=batch.war/subsystem=batch-jberet}} processes each job instance then each job execution of that job instance.
> One possibly helpful option for the web console would be to add a new description attribute to indicate the resource may be slow to process. The web console might be able to run a background task to populate data rather than locking up the UI. There would still be an issue with a large memory footprint here however.
> JBeret might want to consider having a way to archive jobs too rather than just purge them. Some users may want to keep all job execution data. Archiving this data could reduce the size of the current data being retrieved.



--
This message was sent by Atlassian Jira
(v7.13.8#713008)