Stop, Start, and Check Servers

Essbase platform components run as managed servers on a WebLogic application server. Start them all using start.sh or start.cmd in esstools. Stop them all using stop.sh or stop.cmd. You can also list the status, and start or stop individual server components.

For Linux, Essbase can be started using the startup script start.sh. Status can be checked using status.sh, and components can be stopped using stop.sh.

For Windows, Essbase can be started using the startup commands start.cmd. Status can be checked using status.cmd, and components can be stopped using stop.cmd.

These are located in the configuration directory of your domain, under esstools/bin.
  • For Linux:
    <Domain_Home>/esstools/bin
  • For Windows:
    <Domain_Home>\esstools\bin

Check Managed Server Status

Run the status to see which managed servers are running in the Essbase platform:

  • For Linux:
    <Domain_Home>/esstools/bin/status.sh
    In this example, AdminServer and Essbase Server are shown in the script as running:
    [/scratch/user/oracle_home/user_projects/domains/essbase_domain/esstools/bin]$ ./status.sh
    Domain status; Using domainHome:
    /scratch/user/oracle_home/user_projects/domains/essbase_domain ...
    Initializing WebLogic Scripting Tool (WLST) ...
    Welcome to WebLogic Server Administration Scripting Shell
    Type help() for help on available commands
    Reading domain...
    /Servers/AdminServer/ListenPort=7001
    Accessing admin server using URL t3://myhost:7001
    Server Name       Server Status   Type       Essbase Status  Machine                  
     ...........       .............   ....       .............   ...........              
     AdminServer       RUNNING         Server     --              myhost.example.com    
     essbase_server1   RUNNING         Server     Active          myhost.example.com 
     eas_server1       --              Server     --              --                         
    

    If the status fails with an error, for example, that the connection to Node Manager was refused, the log should be analyzed for the cause of the error.

  • For Windows:
    <Domain_Home>\esstools\bin\status.cmd

    Note:

    On your Windows machine, you can also check the status of Windows Services. In Windows Task Manager > Services, you can view Windows services that are running in the background.

Start All Servers

Run startup, without any arguments, to start Node Manager, all managed servers, and AdminServer, in the Essbase platform:

  • For Linux:
    <Domain_Home>/esstools/bin/start.sh
  • For Windows:
    <Domain_Home>\esstools\bin\start.cmd

The startup always starts WebLogic Node Manager. If no argument is specified to the startup script, as in the example above, Node Manager starts AdminServer, and AdminServer starts managed servers, including the Essbase server (and EAS server, if applicable).

Alternately, the startup can be used to start a particular server, if a server name argument is provided:
start.sh -i <server name>
If a server name is specified to the startup script, Node Manager starts the specified server only. For example, to start only the Essbase server,
  • For Linux:
    ./start.sh -i essbase_server1
  • For Windows:
    .\start.cmd -i essbase_server1

Before stopping servers: prepare for shutdown

When you need to stop your Essbase services, consider the impact to active users. Enable/disable commands are available to allow you to prevent new connections, and you can write scripts using wait times to ensure that existing user operations have time to finish. If some operations are running for too long, you can terminate sessions and individual operations. Once all users are off the system, you can proceed to stop the services.

If you need to stop active user sessions, you can use the following workflow:
  1. In the application settings, disable connections and commands (In MaxL, Alter Application disable connects|commands)
  2. In the database settings, disable startup (In MaxL, Alter Database disable startup)
  3. At the Essbase server level, log out all users from the system and terminate all active requests (In MaxL, Alter System logout session and kill request)
  4. For each cube, verify that all sessions and requests are terminated (in MaxL, Display Session)
  5. Shut down all the applications (in MaxL, Alter System unload application)

Stop All Servers

Run the stop (Linux) script or (Windows) command, without any arguments, to stop Node Manager, AdminServer, and all managed servers in the Essbase platform:

  • For Linux:
    <Domain_Home>/esstools/bin/stop.sh
  • For Windows:
    <Domain_Home>\esstools\bin\stop.cmd

If a server name is specified to the stop (Linux) script or (Windows) command, Node Manager stops the specified server. For example, if you installed EAS server and don’t need it running now, you can use:

./stop.sh -i eas_server1

Start Secondary Managed Server when AdminServer is Not Available

When Essbase is configured for high availability, on the secondary (or any additional failover) node, if you need to start the WebLogic managed server when the primary node becomes unavailable, follow these instructions. Provide the managed server name of the secondary node, and the AdminServer URL, even though the AdminServer is unavailable. When prompted, also provide WebLogic administrator credentials.

  1. SSH to the Essbase secondary node (Host 2).

  2. Navigate to <Domain Home>/bin.

  3. Run the script to start the WebLogic managed server, using the following syntax. You need to provide the AdminServer host name and the AdminServer port, even though AdminServer is not available.

    startManagedWebLogic.sh <failover-managed-server-name> t3(s)://<AdminServer-Host>:<AdminServer-Port>

    Linux Example with TLS/SSL:

    ./startManagedWebLogic.sh essbase_server2 t3s://adminserver:7002

    Linux Example without TLS/SSL:

    ./startManagedWebLogic.sh essbase_server2 t3://adminserver:7001

    Windows Example with TLS/SSL:

    .\startManagedWebLogic.cmd essbase_server2 t3s://adminserver:7002

    Windows Example without TLS/SSL:

    .\startManagedWebLogic.cmd essbase_server2 t3://adminserver:7001

Stop Secondary Managed Server when AdminServer is Not Available

When Essbase is configured for high availability, on the secondary (or any additional failover) node, you can stop the WebLogic managed server independently, in case the primary node and AdminServer become unavailable.

  1. SSH to the Essbase secondary node (Host 2).

  2. Navigate to <ORACLE_HOME>/oracle_common/common/bin.

  3. Run the WebLogic Scripting Tool.

    Linux example:

    ./oracle_common/common/bin/wlst.sh

    Windows example:

    .\wlst.cmd
  4. At the wls:/offline> prompt, connect using the following syntax. You need to provide the WebLogic administrator credentials and the secondary (failover) host name and port.

    connect('<weblogic-admin-user>','<weblogic-admin-user-password>','t3(s)://<failover-managedServer-host>:<failover-managedServer-port>')

    Example with TLS/SSL:

    connect('WLadminusername','WLadminpassword','t3s://essbase_server.example.com:9001')

    Example without TLS/SSL:

    connect('WLadminusername','WLadminpassword','t3://essbase_server.example.com:9000')
  5. Shut down the secondary (failover) host.

    Example:

    shutdown('essbase_server2','Server','true',1000, force='true', block='true')
  6. Disconnect from the AdminServer.

    Example:

    disconnect()
  7. Exit the WebLogic Scripting Tool.

    Example:

    exit()

For Windows only - Start and Stop Essbase Service

Windows SCM can be used to start or stop the Essbase service. To do this, open Windows SCM > select Essbase service > select and start or stop the service on Windows SCM.

For Independent (On-Premise) deployments only - rotation of Essbase server logs

For Independent (On-Premise) deployments, log rotation of essbase_server1.out is not supported and you must manually take a backup of the log file, when the servers are stopped. The file is located in <DOMAIN_HOME>/<DOMAIN_NAME>/servers/essbase_server1/logs/.

Backup saves older versions of the log, and should be done as the current log file size becomes very large.