Invision Community 4: SEO, prepare for v5 and dormant account notifications By Matt Monday at 02:04 PM
rahaug Posted November 16, 2016 Posted November 16, 2016 Hi, My server receives random 502 bad gateway almost every hour. I've been trying to figure this one out for almost two weeks but I cannot seem to solve it. Max children isn't reached once since yesterday, but I know of multiple incidents of 502 since yesterday. Do you have any ideas? Any help is deeply appreciated. fpm status report: pool: www process manager: dynamic start time: 15/Nov/2016:06:40:08 -0500 start since: 121517 accepted conn: 349082 listen queue: 0 max listen queue: 0 listen queue len: 0 idle processes: 25 active processes: 2 total processes: 27 max active processes: 8 max children reached: 0 slow requests: 0 I run ubuntu 14.04.5 LTS, nginx 1.4.6, php 7.0.12-1. 4 core Cpu, 8 gb ram. SSD drive.
Feld0 Posted November 21, 2016 Posted November 21, 2016 This sounds like an issue linked to running php-fpm in Unix socket mode rather than in TCP/IP mode (in the latter mode, it'll listen at an IP address and port number like 127.0.0.1:8000). php-fpm handles significant traffic more gracefully in TCP/IP mode. Your syslog, php-fpm, and nginx error logs may have details on the cause, too, if you haven't already checked them. If your php-fpm is configured in socket mode, try changing that to TCP/IP mode. To do that, modify the "listen" setting in the fpm pool configuration (it's probably located in /etc/php/7.0/fpm/pool.d/www.conf) - there should be a comment in the file that explains what to set it to. Remember to also update your nginx configuration to refer to the IP+port after instead of the socket after changing this.
Recommended Posts
Archived
This topic is now archived and is closed to further replies.