Service on Laravel, consisting of bundles of docker-compose (nginx + php-fpm 7.1.2) is raised at t2.medium(two cores), hanging under LoadBalancer (AWS) + AutoScaling (AWS) is called by cron at 5 am. The script takes about an hour and a half. Problem
: during the script, after 5 minutes, AWS shows the service as unhealthy - when you call service (by IP), nginx throws 502.
docker logs web_container
*AWESOME_IP*- - [02/Mar/2018:05:05:26 +0000] "GET / HTTP/1.1" 499 0 "-" "ELB-HealthChecker/2.0" "-"
AWS checks once per minute.
Ie logs in 5 minutes after running cron starts 499. In parallel, even at night, a few points of contact continually with the service. Cron runs successfully.
Locally, such a problem was not observed, naturally without HealthChecker and additional call points.
Both container pingoud among themselves, both individually react, nginx throws the error, and php container allows to call functions in laravel (php artisan tinker etc..), to work with controllers and base.Plantain knee
: restart the app-Container (php).
Tell me how and where to dig to find and fix the problem?
Of course, as a plan for the future, to pull routine in a separate cron service, which runs only for this purpose, but unfortunately at this time there, and the problem I want to understand.
Thank you very much!UPDATE
alright, I dug further, found an article
it made the request, a living container, gives the right answer
Primary script unknown
Status: 404 Not Found
Content-type: text/html; charset=UTF-8
File not found.
and dead does not give anything at all - hangs after 2-3 minutes drops.
So the problem is directly related to php-fpm for php itself works, but the queries can not handle. Therefore, remove the tag docker.
Prompt where to dig further?Logging php-fpm was disabled, is now included, to recreate the error failed, I'll watch and wait depend on...UPDATE 5.03
The server has again "fallen" and in the log is not written...