def log_url(self, job_execution):
job_instance = JobInstance.find_job_instance_by_id(job_execution.job_instance_id)
project = Project.find_project_by_id(job_instance.project_id)
for spider_service_instance in self.spider_service_instances:
if spider_service_instance.server == job_execution.running_on:
return spider_service_instance.log_url(project.project_name, job_instance.spider_name,
job_execution.service_job_execution_id)
def process_finished_jobs(self, job_status, job_execution_dict):
found_jobs = []
for job_execution_info in job_status[SpiderStatus.FINISHED]:
found_jobs.append(job_execution_info['id'])
job_execution = job_execution_dict.get(job_execution_info['id'])
if not job_execution or job_execution.running_status == SpiderStatus.FINISHED:
# the minimum check
continue
job_execution.start_time = job_execution_info['start_time']
job_execution.end_time = job_execution_info['end_time']
job_execution.running_status = SpiderStatus.FINISHED
res = requests.get(self.log_url(job_execution), headers={"Range": "bytes=-4096"})
res.encoding = 'utf8'
match = re.findall(job_execution.RAW_STATS_REGEX, res.text, re.DOTALL)
if not match:
continue
execution_results = match[0]
job_execution.raw_stats = execution_results
job_execution.process_raw_stats()
job_instance = JobInstance.find_job_instance_by_id(job_execution.job_instance_id)
spider_info = SpiderInfo.get_spider_info(job_instance.spider_name, job_instance.project_id)
spider_info.update_spider_info(job_execution.raw_stats)
return found_jobs
def cancel_spider(self, job_execution):
job_instance = JobInstance.find_job_instance_by_id(
job_execution.job_instance_id)
project = Project.find_project_by_id(job_instance.project_id)
for spider_service_instance in self.spider_service_instances:
if spider_service_instance.server == job_execution.running_on:
if spider_service_instance.cancel_spider(
project.project_name,
job_execution.service_job_execution_id):
job_execution.end_time = datetime.datetime.now()
job_execution.running_status = SpiderStatus.CANCELED
db.session.commit()
break
def run_spider_job(job_instance_id):
'''
run spider by scheduler
:param job_instance_id:
:return:
'''
try:
job_instance = JobInstance.find_job_instance_by_id(job_instance_id)
agent.start_spider(job_instance)
app.logger.info(
'[run_spider_job][project:%s][spider_name:%s][job_instance_id:%s]'
% (job_instance.project_id, job_instance.spider_name,
job_instance.id))
except Exception as e:
app.logger.error('[run_spider_job] ' + str(e))
def run_spider_job(job_instance_id):
'''
run spider by scheduler
:param job_instance_id:
:return:
'''
try:
job_instance = JobInstance.find_job_instance_by_id(job_instance_id)
start_tasks = job_instance.start_tasks
"""start_time = datetime.now() - timedelta(minutes=30)
count = JobExecution.query.filter_by(
job_instance_id=job_instance_id,
running_status=SpiderStatus.RUNNING,
).filter(JobExecution.start_time < start_time).count()
if count > 0:
return"""
count = JobExecution.query.filter_by(
job_instance_id=job_instance_id).filter(
JobExecution.running_status.in_(
[SpiderStatus.PENDING, SpiderStatus.RUNNING])).count()
if count >= job_instance.max_start_tasks:
return
slots = job_instance.max_start_tasks - count
if job_instance.start_tasks > slots:
start_tasks = slots
if start_tasks > 0:
i = 0
while i < start_tasks:
agent.start_spider(job_instance)
i += 1
app.logger.info(
'[run_spider_job][project:%s][spider_name:%s][job_instance_id:%s]'
'[start_tasks:%s][i:%s]' %
(job_instance.project_id, job_instance.spider_name,
job_instance.id, start_tasks, i))
except Exception as e:
app.logger.error('[run_spider_job] ' + str(e))