Hello,
I start using php-fpm since a few hours and encounter following
problem when using the Slow log. It says dump failed in 95% of
logentries
php-slow.log
---
Jun 22 23:48:18.679249 pid 2697 (pool slayer1_production)
+++ dump failed
---
php-fpm.log
---
Jun 22 23:48:18.678962 [WARNING] fpm_request_check_timed_out(), line
146: child 2697, script 'script.php' (pool slayer1_production)
executing too slow (5.067665 sec), logging
Jun 22 23:48:18.679032 [NOTICE] fpm_got_signal(), line 48: received
SIGCHLD
Jun 22 23:48:18.679196 [NOTICE] fpm_children_bury(), line 194: child
2697 stopped for tracing
Jun 22 23:48:18.679211 [NOTICE] fpm_php_trace(), line 139: about to
trace 2697
Jun 22 23:48:18.679273 [ERROR] fpm_trace_get_long(), line 69:
ptrace(PT_IO) failed: Bad address (14)
Jun 22 23:48:18.679313 [NOTICE] fpm_php_trace(), line 167: finished
trace of 2697
---
System is FreeBSD 7.2
Php is 5.2.13 installed from Ports
Php is external on 2 servers working via Tcp
On each server, php-fpm config is set to 200 children.
Also from latest lighttpd 1.4x branch i have a lot of
2010-06-22 23:58:13: (mod_fastcgi.c.3104) connection was dropped after
accept() (perhaps the fastcgi process died), write-offset: 0 socket:
tcp:213.131.228.177:1889
2010-06-22 23:58:21: (mod_fastcgi.c.3104) connection was dropped after
accept() (perhaps the fastcgi process died), write-offset: 0 socket:
tcp:213.131.228.177:1889
2010-06-22 23:58:31: (mod_fastcgi.c.3104) connection was dropped after
accept() (perhaps the fastcgi process died), write-offset: 0 socket:
tcp:213.131.228.178:1889
2010-06-22 23:58:34: (mod_fastcgi.c.3104) connection was dropped after
accept() (perhaps the fastcgi process died), write-offset: 0 socket:
tcp:213.131.228.177:1889
any idea what *could* be going wrong?