Jobstore Oracle drops "OperationalError ORA-03114" after adding a job (add_job)

[srPuebla]

Hi,

Libraries version:
APScheduler-3.5.3
SQLAlchemy-1.3.4
cx_Oracle-7.1.3
python-2.7.14

BackgroundScheduler is started (with job_store) and it works successfully (You can add jobs, jobstore works fine, etc) However, when more then 24 hours inactive (no new jobs are added or modified) the following error occurs when adding a new job "OperationalError('(cx_Oracle.OperationalError) ORA-03114: not connected to ORACLE',) "

Code OK:

job_store = SQLAlchemyJobStore(url=URL_JOBSTORE, tablename = TABLE_JOBSTORE)
_job_scheduler = BackgroundScheduler()
_job_scheduler.add_jobstore(job_store)
_job_scheduler.start()
_job_scheduler.add_job(...) #OK

Code Error:

job_store = SQLAlchemyJobStore(url=URL_JOBSTORE, tablename = TABLE_JOBSTORE)
_job_scheduler = BackgroundScheduler()
_job_scheduler.add_jobstore(job_store)
_job_scheduler.start()
#LONG TIME OPERATIONS 24 hours past
_job_scheduler.add_job(...) # drops OperationErrror exception

Error full:
OperationalError: (cx_Oracle.OperationalError) ORA-03114: not connected to ORACLE\n[SQL: INSERT INTO "jobrestore_PRE_CENTRE2_0" (id, next_run_time, job_state) VALUES (:id, :next_run_time, :job_state)] (Background on this error at: http://sqlalche.me/e/e3q8)\n'

I think that Jobstore has lost connection. Cause if you create again the scheduler it works properly. Is there any way to reconnect the Jobstore to the Database?

Thanks

[agronholm]

I suggest that you try passing engine_options={'pool_pre_ping': True} to the job store. That should prevent any errors caused by the database server closing idle connections.

[srPuebla]

Thanks so much for rapid reply. I am going to check on our application!

[srPuebla]

Sorry Agronholm for delay.

Problem has been solved. Thanks so much, if i have other question, i will post!