Uploaded image for project: 'Jira Data Center'
  1. Jira Data Center
  2. JRASERVER-39544

NPE in Upgrade task of JRA-34394: Watch count gets out of sync when cloning issues

XMLWordPrintable

      NOTE: This bug report is for JIRA Server. Using JIRA Cloud? See the corresponding bug report.

      Next code will throw NPE:

      if (!watchesByIssueId.get(issueId).equals(correctWatches))
{
    result.add(new BrokenIssue(issueId, correctWatches));
}

      in case, that there is a watcher linked to an issue, which was already deleted. Because it will be not able to get necessary information from this issue.

      Diagnosis

      The following NullPointerException appears in $JIRA_HOME/log/atlassian-jira.log file:

      2014-08-18 02:08:42,924 localhost-startStop-1 INFO      [atlassian.jira.upgrade.UpgradeManagerImpl] ___ Performing Upgrade ____________________
2014-08-18 02:08:43,251 localhost-startStop-1 INFO      [atlassian.jira.upgrade.UpgradeManagerImpl] Performing Upgrade Task: JRA-34394: Fixing incorrect watch count for cloned issues
2014-08-18 02:08:43,610 localhost-startStop-1 ERROR      [atlassian.jira.upgrade.UpgradeManagerImpl] Exception thrown during upgrade: null
java.lang.NullPointerException
	at com.atlassian.jira.upgrade.tasks.UpgradeTask_Build6331$2.consume(UpgradeTask_Build6331.java:105)
	at com.atlassian.jira.upgrade.tasks.UpgradeTask_Build6331$2.consume(UpgradeTask_Build6331.java:96)
	at com.atlassian.jira.entity.SelectQueryImpl$ExecutionContextImpl.consumeWith(SelectQueryImpl.java:240)
	at com.atlassian.jira.upgrade.tasks.UpgradeTask_Build6331.findBrokenIssues(UpgradeTask_Build6331.java:93)
	at com.atlassian.jira.upgrade.tasks.UpgradeTask_Build6331.doUpgrade(UpgradeTask_Build6331.java:63)
	at com.atlassian.jira.upgrade.UpgradeManagerImpl.doUpgradeTaskSuccess(UpgradeManagerImpl.java:693)
	at com.atlassian.jira.upgrade.UpgradeManagerImpl.runUpgradeTasks(UpgradeManagerImpl.java:542)
	at com.atlassian.jira.upgrade.UpgradeManagerImpl.doUpgrade(UpgradeManagerImpl.java:471)
	at com.atlassian.jira.upgrade.UpgradeManagerImpl.doUpgradeIfNeeded(UpgradeManagerImpl.java:413)
	at com.atlassian.jira.upgrade.UpgradeManagerImpl.doUpgradeIfNeededAndAllowed(UpgradeManagerImpl.java:348)
	at com.atlassian.jira.upgrade.UpgradeLauncher.checkIfUpgradeNeeded(UpgradeLauncher.java:106)
	at com.atlassian.jira.upgrade.UpgradeLauncher.start(UpgradeLauncher.java:54)
	at com.atlassian.jira.startup.ActiveServicesLauncher.start(ActiveServicesLauncher.java:42)
	at com.atlassian.jira.startup.DefaultJiraLauncher$3.run(DefaultJiraLauncher.java:133)
	at com.atlassian.jira.config.database.DatabaseConfigurationManagerImpl.doNowOrEnqueue(DatabaseConfigurationManagerImpl.java:324)
	at com.atlassian.jira.config.database.DatabaseConfigurationManagerImpl.doNowOrWhenDatabaseActivated(DatabaseConfigurationManagerImpl.java:214)
	at com.atlassian.jira.startup.DefaultJiraLauncher.postDbLaunch(DefaultJiraLauncher.java:115)
	at com.atlassian.jira.startup.DefaultJiraLauncher.access$100(DefaultJiraLauncher.java:31)
	at com.atlassian.jira.startup.DefaultJiraLauncher$1.run(DefaultJiraLauncher.java:78)
	at com.atlassian.jira.util.devspeed.JiraDevSpeedTimer.run(JiraDevSpeedTimer.java:34)
	at com.atlassian.jira.startup.DefaultJiraLauncher.start(DefaultJiraLauncher.java:73)
	at com.atlassian.jira.startup.LauncherContextListener.contextInitialized(LauncherContextListener.java:71)  <+5> (StandardContext.java:4939) (StandardContext.java:5434) (LifecycleBase.java:150) (ContainerBase.java:1559) (ContainerBase.java:1549)
	at java.util.concurrent.FutureTask.run(Unknown Source)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
	at java.lang.Thread.run(Unknown Source)

      Workaround

      This workaround involves executing SQL queries on the JIRA database and hence we recommend having/taking a backup of the database either using JIRA XML backup (before the upgrade) or using native database backup tools as a precaution.

      1. Stop JIRA.
      2. From the JIRA database, remove the 'WatchIssue' associations which reference non-existing issues using the following SQL query:
        PostgreSQL 
        delete from "userassociation" where "association_type" = 'WatchIssue' and "sink_node_entity" = 'Issue' and "sink_node_id" not in ( select issue."id" from "jiraissue" issue );

        Microsoft SQL Server
        The following query assumes that you are using the schema jiraschema. If you use a different one, you would need to replace jiraschema with yours - you could find that out from $JIRA_HOME/dbconfig.xml file.

        delete from jiraschema.userassociation where "association_type" = 'WatchIssue' and "sink_node_entity" = 'Issue' and "sink_node_id" not in ( select issue."id" from jiraschema.jiraissue issue );

        MySQL
        Thanks to Leos Junek's comment below

        DELETE FROM userassociation WHERE association_type = 'WatchIssue' AND sink_node_entity = 'Issue' AND sink_node_id NOT IN (SELECT issue.id FROM jiraissue issue);

        Oracle
        Thanks to Robert Heckel's comment below

        delete FROM userassociation where association_type = 'WatchIssue' and sink_node_entity = 'Issue' and sink_node_id not in ( select issue.id from jiraissue issue );
commit;
      3. Start JIRA.

              ohernandez@atlassian.com Oswaldo Hernandez (Inactive)
              sdvorscak StanislavA
              Votes:
              1 Vote for this issue
              Watchers:
              24 Start watching this issue

                Created:
                Updated:
                Resolved: