MRatWork Forum by Mustafa Ramadhan
Sawo Project - Kloxo-MR Discussions => Kloxo-MR Technical Helps => Topic started by: hoangsang on 2013-01-11, 23:37:09
-
Hi all.
My server running: Current Version: 6.5.0.candidate.2012-12-17-01
And:
Webserver: Nginx
Php Branch: php3u
Site: Forum VBB4.
Some time client access website is show error:
@charset "UTF-8";.toolsmenu{background:#e9e9e9 url(http://bachngocsach.com/forum/images...btn_middle.png (http://bachngocsach.com/forum/images...btn_middle.png)) repeat-x ;width:100%;display:block;border:1px solid #cecece;float:left;clear:both;margin:0.5em 0}.toolsmenu div{text-align:right;color:#4e4e4e;width:100%;clear:both;fl oat:left}.toolsmenu h1{font-size:14px;font-weight:bold;margin-top:0.25em;text-align:left;display:inline;float:left;margin-left:10px}.toolsmenu h1 a{color:#4e4e4e}.toolsmenu ul{padding:5px 0;float:right;margin-right:10px}.toolsmenu .nopopupgroup li{float:left;margin-left:10px}.toolsmenu .popupgroup a.popupctrl,.toolsmenu .popupgroup .popupmenu a.popupctrl,.toolsmenu .nopopupgroup li a{font-size:11px;color:#4e4e4e;display:block;_display:inl ine;float:left;clear:right;padding:0.2em 0.4em}.toolsmenu .popupgroup a.popupctrl,.toolsmenu .popupgroup .popupmenu a.popupctrl,.toolsmenu .popupgroup .popupmenu:hover a.popupctrl,.toolsmenu .popupgroup .popupmenu:hover .popupctrl a.popupctrl.active{border:0;color:#4e4e4e;padding-right:15px}.toolsmenu li{position:relative;display:block;clear:right}.to olsmenu .popupgroup ul{text-align:left;float:left}.toolsmenu .popupgroup
....................................................
Server: nginx
Date: Fri, 11 Jan 2013 17:02:05 GMT
Content-Type: text/html; charset=UTF-8
Transfer-Encoding: chunked
Connection: keep-alive
X-Powered-By: PHP/5.3.18
Cache-Control: private
Pragma: private
Set-Cookie: bb_thread_lastview=d2ba7df341c311d3ccc2cceda69d124 51f37f3a4a-2-%7Bi-4075_i-1357922545_i-4073_i-1357908589_%7D; path=/forum/
Content-Encoding: gzip
Cache-Control: public
1ec8
‹??????ÿí½ko$Ç•6ø¹ø?¤J? ?ȶÈ*ïMvQh²¯vßÔdËÖ6{ YUIVª«*Ku!›– ´*0 ƒ cv?b0¬eÁ??ë1l¯?~Q?Å| VŸô'úŸ?óœ™u!û"?btaUeÆåDĉs —ß¼zogïƒûׂæ ?
î?ܾ}k'(?—J?YÚ)•®î]
.............................................
Or Web Server response very slow (server uses about 15% of resources)
Please help me!
Thank you!
-
what plugin is active?
make a dd and ioping test to see the disk performance
-
what plugin is active?
make a dd and ioping test to see the disk performance
No plugin is installed, and HDD is SAS 15k, and running raid 10.
But the problem is that someone users is error and other users is normal (load the website is very fast)
-
Same browser?
-
Same browser?
Usually is IE, Chrome. But if change Webserver to Apache is normal no error :(.
I want Nginx :)
-
I see lots of 304 errors
-
I see lots of 304 errors
can you help me?
-
Do you have some plugin ho interact with your theme?
For what i can see its an incompatibility problem of some plugin with webserver.
-
if you are using php-fpm it needs to tune the config. like the max children, requests etc.
how is the sites traffic and servers hardware?
-
if you are using php-fpm it needs to tune the config. like the max children, requests etc.
how is the sites traffic and servers hardware?
monthly
rx | tx | total | avg. rate
------------------------+-------------+-------------+---------------
Dec '12 40.58 GiB | 375.32 GiB | 415.90 GiB | 1.30 Mbit/s
Jan '13 20.22 GiB | 245.32 GiB | 265.54 GiB | 1.20 Mbit/s
------------------------+-------------+-------------+---------------
estimated 29.18 GiB | 354.03 GiB | 383.21 GiB |
daily
rx | tx | total | avg. rate
------------------------+-------------+-------------+---------------
yesterday 924.33 MiB | 10.97 GiB | 11.87 GiB | 1.15 Mbit/s
today 342.13 MiB | 3.74 GiB | 4.07 GiB | 822.78 kbit/s
------------------------+-------------+-------------+---------------
estimated 711 MiB | 7.78 GiB | 8.47 GiB |
And server hardware:
Chipset: 2 x Intel Quad Core E5504 Xeon CPU, 2.0Ghz
RAM: 4 x 2Gb DDR3 ECC
HDD: SAS 3.5", 15k rpm, RAID 10
-
what is the config you are using?
-
what is the config you are using?
nginx.conf
user nginx;
worker_processes 8;
worker_rlimit_nofile 100000;
pid /var/run/nginx.pid;
events {
worker_connections 2048;
use epoll;
}
http {
add_header Cache-Control public;
server_names_hash_max_size 1024;
server_names_hash_bucket_size 128;
client_header_buffer_size 2k;
client_max_body_size 32M;
include '/etc/nginx/conf.d/*.conf';
}
/etc/php-fpm.d/default.conf
[default]
listen = 127.0.0.1:50048
listen.backlog = -1
listen.allowed_clients = 127.0.0.1
user = apache
group = apache
pm = dynamic
pm.max_children = 5
pm.start_servers = 1
pm.min_spare_servers = 1
pm.max_spare_servers = 2
pm.max_requests = 500
;pm.status_path = /status
;ping.path = /ping
;ping.response = pong
request_terminate_timeout = 0
request_slowlog_timeout = 0
slowlog = /var/log/php-fpm/slow.log
rlimit_files = 1024
rlimit_core = 0
;chroot =
;chdir = /var/www
catch_workers_output = yes
security.limit_extensions = .php .php3 .php4 .php5
env[HOSTNAME] = $HOSTNAME
env[PATH] = /usr/local/bin:/usr/bin:/bin
env[TMP] = /tmp
env[TMPDIR] = /tmp
env[TEMP] = /tmp
;php_admin_value[sendmail_path] = /usr/sbin/sendmail -t -i -f webmaster@domain.com
;php_flag[display_errors] = off
php_admin_value[error_log] = /var/log/php-fpm/error.log
php_admin_value[session.save_path] = /var/lib/php/session
php_admin_flag[log_errors] = on
php_admin_value[memory_limit] = 64M
/home/nginx/conf/globals/php-fpm.conf
### MR -- just enough remove # below for enable nginx cache
#fastcgi_cache fcache;
fastcgi_cache_valid 200 1h;
fastcgi_cache_valid 301 1h;
fastcgi_cache_valid 302 1h;
fastcgi_cache_valid any 1m;
### MR -- must be using nginx with including ngx_purge_cache
### must be using special nginx (nginx with ngx_purge_cache include
location ~ /purge(/.*) {
### MR -- just enough remove # below for enable purge_cache
#proxy_cache_purge pcache "$scheme://$host$1$request_method";
allow 127.0.0.1;
deny all;
}
### MR -- for userdir (access with http://ip/~user
location ~ ^/~([^/]+)/(.+.php)$ {
if (!-f /home/$1/public_html/$2) {
rewrite ^ 404;
}
alias '/home/$1/public_html/$2';
fastcgi_pass 127.0.0.1:$fpmport;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include fastcgi_params;
}
### MR -- for userdir (access with http://ip/~user
location ~ ^/~(.+?)(/.*)?$ {
alias '/home/$1/public_html$2';
autoindex on;
}
location ~ .php$ {
include fastcgi_params;
fastcgi_pass 127.0.0.1:$fpmport;
fastcgi_index index.php;
fastcgi_param REDIRECT_STATUS 200;
fastcgi_split_path_info ^(.+.php)(/.+)$;
fastcgi_param SCRIPT_FILENAME $rootdir$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_script_name;
fastcgi_param PHP_ADMIN_VALUE 'sendmail_path=/usr/sbin/sendmail -t -i';
fastcgi_param PHP_ADMIN_VALUE 'sendmail_from=$domain';
}
location / {
limit_conn addr 25;
## MR - increasing upload size to 15M
client_max_body_size 15M;
client_body_buffer_size 128k;
try_files $uri $uri/ /index.php;
}
-
depending on your server configuration, you should consider increasing the values. specially the php-fpm one.
-
depending on your server configuration, you should consider increasing the values. specially the php-fpm one.
You can specific value examples?
-
depending on your server configuration, you should consider increasing the values. specially the php-fpm one.
You can specific value examples?
that will depend on circumstances, but you can first increase the maxchildren to something 12/15, 5 is really not good for you.
-
depending on your server configuration, you should consider increasing the values. specially the php-fpm one.
You can specific value examples?
that will depend on circumstances, but you can first increase the maxchildren to something 12/15, 5 is really not good for you.
I'm changed to:
pm.max_children = 70
pm.start_servers = 50
pm.min_spare_servers = 40
pm.max_spare_servers = 55
pm.max_requests = 1000
But access by Chrome is very slow :(
-
70 is a too large value. just set something between 12 to 15 depending on traffic.
what the log files is saying? specially the slowlog?
any errors, timeouts, or memory overload?
-
Calculate by memory usage too. like if one php-fpm child process is taking 150 MB ram then 70 process will take 10,500 MB which will end up giving you 500 error and probably locking down the server. so choose wisely by using your brain. Good Luck.
-
Calculate by memory usage too. like if one php-fpm child process is taking 150 MB ram then 70 process will take 10,500 MB which will end up giving you 500 error and probably locking down the server. so choose wisely by using your brain. Good Luck.
Thanks. This my log
[22-Jan-2013 17:07:21] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 17:08:22] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 19:31:35] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 19:46:54] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 19:51:59] ERROR: unable to read what child say: Bad file descriptor (9)
[22-Jan-2013 20:46:03] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 21:24:33] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 21:47:16] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 21:59:59] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 22:07:46] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 22:07:53] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 22:29:44] ERROR: unable to read what child say: Bad file descriptor (9)
[22-Jan-2013 22:57:50] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[22-Jan-2013 23:25:17] WARNING: [pool bachngocsach] server reached pm.max_children setting (5), consider raising it
[23-Jan-2013 00:56:58] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:56:58] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:08] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:41] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 00:59:41] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 01:04:19] ERROR: unable to read what child say: Bad file descriptor (9)
[23-Jan-2013 01:04:19] ERROR: unable to read what child say: Bad file descriptor (9)
ERROR: unable to read what child say: Bad file descriptor (9)
hichic :(
-
Calculate by memory usage too. like if one php-fpm child process is taking 150 MB ram then 70 process will take 10,500 MB which will end up giving you 500 error and probably locking down the server. so choose wisely by using your brain. Good Luck.
I'm changed to:
pm.max_children = 15
pm.start_servers = 14
pm.min_spare_servers = 13
pm.max_spare_servers = 15
pm.max_requests = 1000
And view log:
[23-Jan-2013 01:45:30] WARNING: [pool bachngocsach] server reached pm.max_children setting (15), consider raising it
-
You need to configure php mysql nginx for your settings this mean you have to know how much ram and how many cpu do you have
Normally i configure mysql to take at maximum 50% (mysqltuner.pl)of the memory and let other for caching and web server.
Start from this and observe were you have to modify settings.
-
As i told, you should have to raise the max children. which log file it is? /var/log/php-fpm/error.log ?
looks like segfault or open files overload.
increase the connection and children.
-
Calculate by memory usage too. like if one php-fpm child process is taking 150 MB ram then 70 process will take 10,500 MB which will end up giving you 500 error and probably locking down the server. so choose wisely by using your brain. Good Luck.
I'm changed to:
pm.max_children = 15
pm.start_servers = 14
pm.min_spare_servers = 13
pm.max_spare_servers = 15
pm.max_requests = 1000
And view log:
[23-Jan-2013 01:45:30] WARNING: [pool bachngocsach] server reached pm.max_children setting (15), consider raising it
Set this to:
pm.max_children = 30
pm.start_servers = 2
pm.min_spare_servers = 2
pm.max_spare_servers = 6
pm.max_requests = 1000