Page History

...

      run_protege<PROTÉGÉ INSTANCE>.sh
 

•  Navigate to the Explanation.Server directory and transfer the following files:

...

      cd /usr/local/protege/Protégé_x.x/<PROTÉGÉ INSTANCE>/Explanation.Server-x.x.x

• Open the start_explanation_server.sh file, and modify the following line to match the newest Protégé server version number:

      protege_install=/usr/local/protégé/Protégé_x.x/<PROTÉGÉ INSTANCE>/Protege.Server-x.x.x

• If there is a previous version of the explanation server running, shutdown is required before starting the new instance. To verify, type 'ps -ef | grep server' to  view all running server instances.  If an explanation server is running, you should see two different processes, one for the explanation server, as well as the Protege server: Image Removed.

• Navigate to the previous explanation server folder version (this folder should be located in the same directory as the newest version), and run the explanation server folder shutdown script (./stop_explanation_server.sh). You can type 'ps -ef | grep server' to verify that the explanation server process has been terminated:

•  Navigate to the new explanation server folder (usr/local/protégé/Protégé_x.x/<PROTÉGÉ INSTANCE>/Explanation.Server-x.x.x) and run './start_explanation_server.sh -p <port> -url jdbc:mysql://<database server address>:<port>/<database name> -t <tablename> -U  <db_username> -P <db_password>'.  --protege-standalone --port <port> /app/protege/data/Protege_x.x/<PROTEGE-INSTANCE>/database-project.pprj.  To determine the proper port number to use, refer to the tier specific info page. After the server has initiated startup, type 'tail -f nohup.out' to view start up status.  Within the nohup.out file,  The explanation server indicates  it is started when you can see "Jena, Classification, and Extracting' all have been initialized, and finally a "Server started, listening on port xxxx ....."  message.

• After the explanation server is up and running, the same steps must be performed for proper Protege server startup.  If there is a previous version running, it must be shutdown before starting the new version. Type 'ps -ef | grep server' to view all server instances, and if a previous instance is running, navigate to the previous version's server folder (/usr/local/protege/Protege_x.x/<PROTEGE_INSTANCE>/<previous Protege server instance>, and run the server shutdown script (./shutdown_protege_server.sh localhost:<port>). Notice for the Protege server shutdown script the port is passed as an argument. This will be needed for proper server shutdown.

• Before restarting the new Protege server instance, verify that the RMI registry is running on the server. To verify, type ps -ef | grep 'rmi' to view all active RMI registries.  If you do not see the port number which the server is listening for, then start the rmi registry by running 'run_rmiregistry.sh'.  Type 'ps -ef  | grep rmiregistry' to verify that it's back up.  This step is not required to be performed all the time, when shutting down previous instances to restart with a new instance, the RMI registry can usually be left running. It may be a good idea to shutdown the registry and restart to refresh the process.

• Navigate to your Protégé server directory, and type 'ln -s  /app/protégé/data/luceneworkflow' .  This command will create a soft link to the lucene workflow folder which is centralized in the specified directory.  Do the same for the workflow folder, if configuring a BiomedGT instance of Protégé:  'ln -s /app/protégé/data/workflow'.     

• Restart the server by running 'run_protege_server_nci.sh.'  Type 'tail -f console.txt'  to verify that it's up and running.  You will see a message displaying "Protege server ready to accept connections".   

...