hemos hecho una instalación básica de un suiteCRM 8.5 y nos encontramos que puntualmente se lanza esta query:
SELECT id FROM job_queue WHERE execute_time <= ‘2024-02-21 20:56:01’ AND status = ‘queued’ ORDER BY
(No se si falta parte de la query, pero no me muestra más el processlist en la terminal)
y a partir de ahí, empieza a, cada minuto, no terminar la siguiente query:
SELECT id FROM job_queue WHERE status=‘running’ AND date_modified <= ‘2024-02-20 20:57:01’
hasta que bloquea el sistema por acumulación de conexiones activas. Luego, a pesar de matar las querys, se quedan en estado Killed y no reaccionan.
La instalación consta de lo siguiente:
Ubuntu 22.04.3 LTS
SuiteCRM8.5
PHP 8.2.15
10.6.16-MariaDB
¿Alguien sabe qué puede estar pasando? No hay nada configurado ni ningún documento en la base de datos, es una instal·lación de zero que no ha sido tocada todavia.
Hello, the cron job is configured exactly how it’s told by SuiteCRM itself after the instalation, and this is the only cron job that exists in the machine:
cd /var/www/html/suitecrm/public/legacy; php -f cron.php > /dev/null 2>&1
Not really, I check the logs files and the only I get is the error that the system shows when it can’t create a new connection to the Database because it’s already full.
Thu Feb 22 07:32:51 2024 [67434][1][FATAL] Mysqli_query failed.
Thu Feb 22 07:32:51 2024 [67434][1][FATAL] Query Failed: SELECT id FROM job_queue WHERE status=‘done’ AND resolution=‘success’ AND date_modified <= ‘2024-01-22 23:26:01’: MySQL error 2006: MySQL server has gone away
Ok, but that is a sign that your DB server is not ok, and that is enough of a problem to cause all sorts of symptoms in SuiteCRM. You definitely need to solve that FATAL before continuing.
Yeah, this is when the DB was full of the connections mentioned before. Before getting full, this error never appeared, it just start appearing after it, when, due to the situation, it was impossible to connect to the DB server. This FATAL is a consecuence of the problem mentioned at the start of the post.
I ended restarting the machine, now the cron job is working fine, the CRM is working, the BD is working. It happened before, as this is not yet a production environment, there is no real trouble with the situation and yesterday, after a week working fine, happened this again, out of no reason. So yes, this is very weird, and I am uncapable to find any cause of it.
Thanks for the answer, maybe I can use the timeout to stop the remaining connections, but this will not solve the problem itself, maybe it will solve, partially, the consecuence, but the cause will still be there, and the cron job will stop being done, which is definetly not desired.