RO_management

Home > Management > RO

---

Disclaimer: The following management manual assumes that configuration of the whole software has been done according to the configuration manual.

Troubleshooting

Managing consistency

Cleaning up the database

RO retrieves information based on the resources advertised by its managed RMs. As these resources may change in a future (some new may be added, others may be removed), there is sometimes necessary to clean up the RO database and restart it.

If you find a message like ObjectId(...) not found in a log or as an output returned by a GENIv3 method; please stop the RO, clean its database and restart it:

1. Execute special command in initscript:

   /etc/init.d/felix-ro force-reload
   

If this did not work or is not available in your deployment, do the following:

1. Stop RO bykill -9 $PID on the correct process(es)

2. Enter RO database. In a console:

   $ mongo
   root@dc1-ro:~# mongo
   MongoDB shell version: 2.0.6
   connecting to: test
   > show dbs
   felix_ro	0.203125GB
   local	(empty)
   test	(empty)
   > use felix_ro
   switched to db felix_ro
   > db.dropDatabase()
   

3. Add the peers to RO (only once after cleaning!):

   # Manually inserting data for managed RMs
   cd /opt/felix/resource-orchestrator/modules/resource/orchestrator/deploy
   ./configure_rms.sh
   
   # OR
   
   # Running a script where the entries are already configured (create one if it does not exist)
   cd /opt/felix/resource-orchestrator/modules/resource/orchestrator/src/admin/db
   ./add_local_rms.sh
   

4. Restart RO:

   cd /opt/felix/resource-orchestrator/modules/resource/orchestrator/src
   
   # Use init script
   /etc/init.d/felix-ro start
   
   # OR
   
   # Send process to background
   python main.py > /dev/null 1>&2 &