This script is executed by job scheduler Azkaban every day and have been working fine for more than a week now. But yesterday it failed unexpectedly. No exception was thrown, no fatal error and no error log entry. I am inserting a count of records processed to database from destruct function that worked fine but shows that only 241 of 7060 records where processed.
Where I am at now:
- If district was called that means no uncaught exception, no fatal error.
- 241 records processed to script was working fine.
- no error in php error log, syslog, logs of azkaban
- script stopped after 50 seconds. normally it takes 30+ minutes. so not timeout.
- it is unlikely to exceed Memory limit. only fetching and processing one record at a time.
Is it possible OS or azkaban or some other process stopped my script, if so how can I find out?
System info : ubuntu server with 2GB of free ram. And 30 gb free memory. CPU usage was also normal at the time of failure.
With no error in logs I am totally confused.