[FIX] procurement: avoid to run multiple scheduler

Avoid to run the procurement scheduler multiple times in the same time, otherwise, they tend to interfere with each others
This commit is contained in:
Denis Ledoux 2014-07-18 13:56:42 +02:00
parent 738270557f
commit 22bbfdaab1
1 changed files with 14 additions and 1 deletions

View File

@ -19,11 +19,14 @@
#
##############################################################################
import logging
import threading
from openerp import pooler
from openerp import pooler, SUPERUSER_ID, tools
from openerp.osv import fields, osv
_logger = logging.getLogger(__name__)
class procurement_compute_all(osv.osv_memory):
_name = 'procurement.order.compute.all'
_description = 'Compute all schedulers'
@ -47,6 +50,16 @@ class procurement_compute_all(osv.osv_memory):
proc_obj = self.pool.get('procurement.order')
#As this function is in a new thread, i need to open a new cursor, because the old one may be closed
new_cr = pooler.get_db(cr.dbname).cursor()
scheduler_cron_id = self.pool['ir.model.data'].get_object_reference(new_cr, SUPERUSER_ID, 'procurement', 'ir_cron_scheduler_action')[1]
# Avoid to run the scheduler multiple times in the same time
try:
with tools.mute_logger('openerp.sql_db'):
new_cr.execute("SELECT id FROM ir_cron WHERE id = %s FOR UPDATE NOWAIT", (scheduler_cron_id,))
except Exception:
_logger.info('Attempt to run procurement scheduler aborted, as already running')
new_cr.rollback()
new_cr.close()
return {}
for proc in self.browse(new_cr, uid, ids, context=context):
proc_obj.run_scheduler(new_cr, uid, automatic=proc.automatic, use_new_cursor=new_cr.dbname,\
context=context)