If you want to use a unique cron job for all subscriptions system then follow the following code in your subscription module in .py file.

 def set_process(self, cr, uid, ids, context=None):
        for row in self.read(cr, uid, ids, context=context):
            mapping = {'name':'name','interval_number':'interval_number','interval_type':'interval_type','exec_init':'numbercall','date_init':'nextcall'}
            res = {'model':'subscription.subscription', 'args': repr([[row['id']]]), 'function':'model_copy', 'priority':6, 'user_id':row['user_id'] and row['user_id'][0]}
            for key,value in mapping.items():
                res[value] = row[key]
            id = self.pool.get('ir.cron').create(cr, uid, res)
            self.write(cr, uid, [row['id']], {'cron_id':id, 'state':'running'})
        return True

    def model_copy(self, cr, uid, ids, context=None):
        for row in self.read(cr, uid, ids, context=context):
            if not row.get('cron_id',False):
                continue
            cron_ids = [row['cron_id'][0]]
            remaining = self.pool.get('ir.cron').read(cr, uid, cron_ids, ['numbercall'])[0]['numbercall']
            try:
                (model_name, id) = row['doc_source'].split(',')
                id = int(id)
                model = self.pool.get(model_name)
            except:
                raise osv.except_osv(_('Wrong Source Document!'), _('Please provide another source document.\nThis one does not exist!'))

            default = {'state':'draft'}
            doc_obj = self.pool.get('subscription.document')
            document_ids = doc_obj.search(cr, uid, [('model.model','=',model_name)])
            doc = doc_obj.browse(cr, uid, document_ids)[0]
            for f in doc.field_ids:
                if f.value=='date':
                    value = time.strftime('%Y-%m-%d')
                else:
                    value = False
                default[f.field.name] = value

            state = 'running'

            # if there was only one remaining document to generate
            # the subscription is over and we mark it as being done
            if remaining == 1:
                state = 'done'
            id = self.pool.get(model_name).copy(cr, uid, id, default, context)
            self.pool.get('subscription.subscription.history').create(cr, uid, {'subscription_id': row['id'], 'date':time.strftime('%Y-%m-%d %H:%M:%S'), 'document_id': model_name+','+str(id)})
            self.write(cr, uid, [row['id']], {'state':state})
        return True

    def unlink(self, cr, uid, ids, context=None):
        for record in self.browse(cr, uid, ids, context or {}):
            if record.state=="running":
                raise osv.except_osv(_('Error!'),_('You cannot delete an active subscription!'))
        return super(subscription_subscription, self).unlink(cr, uid, ids, context)

    def set_done(self, cr, uid, ids, context=None):
        res = self.read(cr,uid, ids, ['cron_id'])
        ids2 = [x['cron_id'][0] for x in res if x['id']]
        self.pool.get('ir.cron').write(cr, uid, ids2, {'active':False})
        self.write(cr, uid, ids, {'state':'done'})
        return True

    def set_draft(self, cr, uid, ids, context=None):
        self.write(cr, uid, ids, {'state':'draft'})
        return True
subscription_subscription()

class subscription_subscription_history(osv.osv):
    _name = "subscription.subscription.history"
    _description = "Subscription history"
    _rec_name = 'date'
    _columns = {
        'date': fields.datetime('Date'),
        'subscription_id': fields.many2one('subscription.subscription', 'Subscription', ondelete='cascade'),
        'document_id': fields.reference('Source Document', required=True, selection=_get_document_types, size=128),
    }
subscription_subscription_history()