So just out of nowhere, I found that my Nextcloud instance was running REALLY slow. The webpage is basically inaccessible. I eventually get a timeout and a 502 error, and looking at the logs, it looks like php-fpm is to blame, but I can't understand why all of a sudden it would run out of server and children processes when there are barely any connections happening at a given time. Here is the error that comes up when a connection is trying to be processed:
Could there be a php bug maybe? I don't know where to begin to look to understand why there is a need to spin up so many threads and still have a failed connection.
This is FreeBSD 11.1, in an iocage jail on a mirrored zfs pool
Code:
[12-Feb-2018 19:13:59] WARNING: [pool www] seems busy (you may need to increase pm.start_servers, or pm.min/max_spare_servers), spawning 8 children, there are 0 idle, and 16 total children
[12-Feb-2018 19:14:00] WARNING: [pool www] seems busy (you may need to increase pm.start_servers, or pm.min/max_spare_servers), spawning 16 children, there are 0 idle, and 18 total children
[12-Feb-2018 19:14:01] WARNING: [pool www] server reached pm.max_children setting (20), consider raising it
Could there be a php bug maybe? I don't know where to begin to look to understand why there is a need to spin up so many threads and still have a failed connection.
This is FreeBSD 11.1, in an iocage jail on a mirrored zfs pool