Update Documentation about Adding Node Name in a Crowd Cluster

XMLWordPrintable

    • Type: Suggestion
    • Resolution: Unresolved
    • None
    • Component/s: Documentation
    • None
    • 3

      https://confluence.atlassian.com/crowd/moving-to-crowd-data-center-935372358.html
      In the Adding node names section of the documentation above it suggest using the following example:

      CATALINA_OPTS=-Dcluster.node.name=node-1
      

      However, as of the latest Crowd 3.2.1, it is still using JAVA_OPTS for the system properties:

      setenv.sh
      JAVA_OPTS="-Xms128m -Xmx512m -Dfile.encoding=UTF-8 $JAVA_OPTS"
      
      export JAVA_OPTS
      
      # set the location of the pid file
      if [ -z "$CATALINA_PID" ] ; then
          if [ -n "$CATALINA_BASE" ] ; then
              CATALINA_PID="$CATALINA_BASE"/work/catalina.pid
          elif [ -n "$CATALINA_HOME" ] ; then
              CATALINA_PID="$CATALINA_HOME"/work/catalina.pid
          fi
      fi
      export CATALINA_PID
      
      setenv.bat
      set JAVA_OPTS=%JAVA_OPTS% -Xms128m -Xmx512m -Dfile.encoding=UTF-8
      
      rem Checks if the JAVA_HOME has a space in it (can cause issues)
      SET _marker=%JAVA_HOME: =%
      IF NOT "%_marker%" == "%JAVA_HOME%" ECHO JAVA_HOME "%JAVA_HOME%" contains spaces. Please change to a location without spaces if this causes problems.
      

      Suggestion

      Suggest to change the documentation from:

      CATALINA_OPTS=-Dcluster.node.name=node-1
      

      to:

      JAVA_OPTS=-Dcluster.node.name=node-1
      

              Assignee:
              Wazza
              Reporter:
              Damien Tan
              Votes:
              2 Vote for this issue
              Watchers:
              5 Start watching this issue

                Created:
                Updated: