Re: fpm problem
On 07/26/2018 04:56 PM, Gokan Atmaca wrote:
>>
>> My advice is :
>>
>> - try to use TCP/IP for communication between FPM and NginX
>
> I read that UNIX socket is better.
Unix socket is faster than TCP/IP but in your case the OS may require
additional tuning. Therefor I proposed TCP/IP.
>
> On Thu, Jul 26, 2018 at 4:05 PM, Georgi Naplatanov wrote:
>> On 07/26/2018 03:37 PM, Gokan Atmaca wrote:
Can you attach your PHP-FPM configuration file and NginX virtual host
section for the site?
>>>
>>> Hello
>>>
>>> Debian stable (strech).
>>>
>>> packages:
>>> php7.2-fpm
>>> 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
>>> server-side, HTML-embedded scripting language (FPM-CGI binary)
>>> nginx 1.10.3-1+deb9u1
>>> all small, powerful, scalable web/proxy server
>>>
>>> Fpm config:
>>>
>>> [www]
>>> user = www-data
>>> group = www-data
>>> listen = /run/php/php7.2-fpm.sock
>>> listen.owner = www-data
>>> listen.group = www-data
>>> pm = dynamic
>>> pm.max_children = 5000
>>> pm.start_servers = 4000
>>> pm.min_spare_servers = 4000
>>> pm.max_spare_servers = 50> --- Nginx config:
>>
>> The configuration is set for very high load. You've used Unix sockets.
>>
>> My advice is :
>>
>> - try to use TCP/IP for communication between FPM and NginX
>>
>> Kind regards
>> Georgi
>>
>>>
>>> server {
>>> listen 80;
>>>
>>> root /home/website/;
>>> index index.php index.html index.htm;
>>> client_max_body_size 256M;
>>> server_name y.br
>>>
>>>
>>>
>>> location = /favicon.ico {
>>> log_not_found off;
>>> access_log off;
>>>}
>>>
>>>
>>> if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
>>> return 403;
>>> }
>>>
>>>
>>> location ~*
>>> \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$ {
>>> expires 365d;
>>> add_header Cache-Control "public";
>>>}
>>>
>>>
>>> location ~* \.(?:css|js)$ {
>>> expires 1y;
>>> access_log off;
>>> add_header Cache-Control "public";
>>> }
>>>
>>>
>>> location / {
>>> try_files $uri $uri/ /index.php?q=$uri&$args;
>>>
>>>
>>> }
>>>
>>> location /my{
>>> allow x.x.x.x;
>>> allow x.x.x.x;
>>> if ($scheme = http) { return 301 https://$host$request_uri;}
>>> deny all;
>>> error_page 403 error403.html;
>>> try_files $uri $uri/ /index.php?q=$uri&$args;
>>> }
>>>
>>> location ~ \.php$ {
>>> try_files $uri =404;
>>> fastcgi_split_path_info ^(.+\.php)(/.+)$;
>>> fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
>>> fastcgi_index index.php;
>>> fastcgi_param SCRIPT_FILENAME $request_filename;
>>> include fastcgi_params;
>>> }
>>>
>>> gzip on;
>>> gzip_disable "msie6";
>>> gzip_comp_level 9;
>>> gzip_min_length 1100;
>>> gzip_buffers 16 8k;
>>> gzip_proxied any;
>>> gzip_types
>>> text/plain
>>> text/css
>>> text/js
>>> text/xml
>>> text/javascript
>>> application/javascript
>>> application/x-javascript
>>> application/json
>>> application/xml
>>> application/xml+rss;
>>>
>>> On Thu, Jul 26, 2018 at 2:56 PM, Georgi Naplatanov wrote:
On 07/26/2018 12:45 PM, Gokan Atmaca wrote:
>> What happened after you did what the warning message suggested to do?
>
> 502 gateway timeout
Can you attach your PHP-FPM configuration file and NginX virtual host
section for the site?
Which Debian branch are you using - stable, testing or unstable? Where
did you get PHP 7.2 packages from ?
Kind regards
Georgi
> On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber
> wrote:
>> Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST:
>>
>>> we are receiving an error as below. what is the cause of the problem?
>>> (32g ram , 240g ssd , xeon 8 cor cpu)
>>>
>>> WARNING: [pool www] seems busy (you may need to increase
>>> pm.start_servers, or pm.min/max_spare_servers), spawning 32 children,
>>> there are 1 idle, and 143 total children
>>
>> What happened after you did what the warning message suggested to do?
>>
>> --
>> Regards
>> mk
>>
>
>>
Re: fpm problem
Am Donnerstag, den 26.07.2018, 16:38 +0300 schrieb Gokan Atmaca: > > Could it be that starting 4000 fpm processes eats up all your > > resources > > and starting the fpm-master process fails? Can you please check if > > the > > php7.2-fpm service is running? > > Yes , services running... > > > Can you try something like this? > > > > pm = dynamic > > pm.max_children = 250 > > pm.start_servers = 25 > > pm.min_spare_servers = 15 > > pm.max_spare_servers = 30 > > fpm, 8 mb per transaction. 32g ram... I have 32G RAM. I just want to > use 24g ram. > > 24576/8 = 3072 (pm.max.child) > is this logic wrong? > I think that php-fpm needs more than 8 MB RAM per process. On my machine they currently around 50 MB per process. And you do not have the guarantee that they will not need more than this. In your first mail the warning says that you are currently using 143 child processes. Why do you want to start 4000 at the beginning? If you do not want to have a dynamic adjustment of the fpm processes, this is what 'pm = dynamic' does, that you can try 'pm = static'. > > S UID PID PPID C PRI NI RSSSZ WCHAN TTY TIME > CMD > S33 9343 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9344 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9345 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9347 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9348 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9349 594 0 80 0 5336 40706 SyS_ep ?00:00:00 > nginx > S33 9346 594 0 80 0 6324 40706 SyS_ep ?00:00:00 > nginx > S33 9342 594 0 80 0 7004 40772 SyS_ep ?00:00:00 > nginx > S33 4828 4792 0 80 0 7400 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4829 4792 0 80 0 7400 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4830 4792 0 80 0 7400 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4831 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4832 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4833 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4834 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4835 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4836 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4837 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4838 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4839 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4840 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4841 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4842 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4843 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4844 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4845 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4846 4792 0 80 0 7404 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4847 4792 0 80 0 7408 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4848 4792 0 80 0 7408 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4849 4792 0 80 0 7408 87637 skb_wa ?00:00:00 > php-fpm7.2 > S33 4850 4792 0 80 0 7408 87637 skb_wa ?00:00:00 > php-fpm7.2 > > > On Thu, Jul 26, 2018 at 4:00 PM, Jens Sauer > wrote: > > Am Donnerstag, den 26.07.2018, 15:37 +0300 schrieb Gokan Atmaca: > > > > Can you attach your PHP-FPM configuration file and NginX > > > > virtual > > > > host > > > > section for the site? > > > > > > Hello > > > > > > Debian stable (strech). > > > > > > packages: > > > php7.2-fpm > > > 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64 > > > server-side, HTML-embedded scripting language (FPM-CGI binary) > > > nginx 1.10.3-1+deb9u1 > > > all small, powerful, scalable web/proxy > > > server > > > > > > > This php7.2-fpm package is definitely not from the stable > > repository. > > php7.2 is not available in stretch and the latest php7.0-fpm > > version > > is 7.0.30-0+deb9u1 > > > > > Fpm config: > > > > > > [www] > > > user = www-data > > > group = www-data > > > listen = /run/php/php7.2-fpm.sock > > > listen.owner = www-data > > > listen.group = www-data > > > pm = dynamic > > > pm.max_children = 5000 > > > pm.start_servers = 4000 > > > pm.min_spare_servers = 4000 > > > pm.max_spare_servers = 5000 > > > > Could it be that starting 4000 fpm processes eats up all your > > resources > > and starting the fpm-master process fails? Can you please check if > > the > >
Re: fpm problem
>
> My advice is :
>
> - try to use TCP/IP for communication between FPM and NginX
I read that UNIX socket is better.
On Thu, Jul 26, 2018 at 4:05 PM, Georgi Naplatanov wrote:
> On 07/26/2018 03:37 PM, Gokan Atmaca wrote:
>>> Can you attach your PHP-FPM configuration file and NginX virtual host
>>> section for the site?
>>
>> Hello
>>
>> Debian stable (strech).
>>
>> packages:
>> php7.2-fpm
>> 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
>> server-side, HTML-embedded scripting language (FPM-CGI binary)
>> nginx 1.10.3-1+deb9u1
>> all small, powerful, scalable web/proxy server
>>
>> Fpm config:
>>
>> [www]
>> user = www-data
>> group = www-data
>> listen = /run/php/php7.2-fpm.sock
>> listen.owner = www-data
>> listen.group = www-data
>> pm = dynamic
>> pm.max_children = 5000
>> pm.start_servers = 4000
>> pm.min_spare_servers = 4000
>> pm.max_spare_servers = 50> --- Nginx config:
>
> The configuration is set for very high load. You've used Unix sockets.
>
> My advice is :
>
> - try to use TCP/IP for communication between FPM and NginX
>
> Kind regards
> Georgi
>
>>
>> server {
>> listen 80;
>>
>> root /home/website/;
>> index index.php index.html index.htm;
>> client_max_body_size 256M;
>> server_name y.br
>>
>>
>>
>> location = /favicon.ico {
>> log_not_found off;
>> access_log off;
>>}
>>
>>
>> if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
>> return 403;
>> }
>>
>>
>> location ~* \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$
>> {
>> expires 365d;
>> add_header Cache-Control "public";
>>}
>>
>>
>> location ~* \.(?:css|js)$ {
>> expires 1y;
>> access_log off;
>> add_header Cache-Control "public";
>> }
>>
>>
>> location / {
>> try_files $uri $uri/ /index.php?q=$uri&$args;
>>
>>
>> }
>>
>> location /my{
>> allow x.x.x.x;
>> allow x.x.x.x;
>> if ($scheme = http) { return 301 https://$host$request_uri;}
>> deny all;
>> error_page 403 error403.html;
>> try_files $uri $uri/ /index.php?q=$uri&$args;
>> }
>>
>> location ~ \.php$ {
>> try_files $uri =404;
>> fastcgi_split_path_info ^(.+\.php)(/.+)$;
>> fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
>> fastcgi_index index.php;
>> fastcgi_param SCRIPT_FILENAME $request_filename;
>> include fastcgi_params;
>> }
>>
>> gzip on;
>> gzip_disable "msie6";
>> gzip_comp_level 9;
>> gzip_min_length 1100;
>> gzip_buffers 16 8k;
>> gzip_proxied any;
>> gzip_types
>> text/plain
>> text/css
>> text/js
>> text/xml
>> text/javascript
>> application/javascript
>> application/x-javascript
>> application/json
>> application/xml
>> application/xml+rss;
>>
>> On Thu, Jul 26, 2018 at 2:56 PM, Georgi Naplatanov wrote:
>>> On 07/26/2018 12:45 PM, Gokan Atmaca wrote:
> What happened after you did what the warning message suggested to do?
502 gateway timeout
>>>
>>> Can you attach your PHP-FPM configuration file and NginX virtual host
>>> section for the site?
>>>
>>> Which Debian branch are you using - stable, testing or unstable? Where
>>> did you get PHP 7.2 packages from ?
>>>
>>> Kind regards
>>> Georgi
>>>
On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber
wrote:
> Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST:
>
>> we are receiving an error as below. what is the cause of the problem?
>> (32g ram , 240g ssd , xeon 8 cor cpu)
>>
>> WARNING: [pool www] seems busy (you may need to increase
>> pm.start_servers, or pm.min/max_spare_servers), spawning 32 children,
>> there are 1 idle, and 143 total children
>
> What happened after you did what the warning message suggested to do?
>
> --
> Regards
> mk
>
>>>
>
Re: fpm problem
> Could it be that starting 4000 fpm processes eats up all your resources
> and starting the fpm-master process fails? Can you please check if the
> php7.2-fpm service is running?
Yes , services running...
> Can you try something like this?
>
> pm = dynamic
> pm.max_children = 250
> pm.start_servers = 25
> pm.min_spare_servers = 15
> pm.max_spare_servers = 30
fpm, 8 mb per transaction. 32g ram... I have 32G RAM. I just want to
use 24g ram.
24576/8 = 3072 (pm.max.child)
is this logic wrong?
S UID PID PPID C PRI NI RSSSZ WCHAN TTY TIME CMD
S33 9343 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9344 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9345 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9347 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9348 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9349 594 0 80 0 5336 40706 SyS_ep ?00:00:00 nginx
S33 9346 594 0 80 0 6324 40706 SyS_ep ?00:00:00 nginx
S33 9342 594 0 80 0 7004 40772 SyS_ep ?00:00:00 nginx
S33 4828 4792 0 80 0 7400 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4829 4792 0 80 0 7400 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4830 4792 0 80 0 7400 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4831 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4832 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4833 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4834 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4835 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4836 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4837 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4838 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4839 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4840 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4841 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4842 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4843 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4844 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4845 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4846 4792 0 80 0 7404 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4847 4792 0 80 0 7408 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4848 4792 0 80 0 7408 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4849 4792 0 80 0 7408 87637 skb_wa ?00:00:00 php-fpm7.2
S33 4850 4792 0 80 0 7408 87637 skb_wa ?00:00:00 php-fpm7.2
On Thu, Jul 26, 2018 at 4:00 PM, Jens Sauer wrote:
> Am Donnerstag, den 26.07.2018, 15:37 +0300 schrieb Gokan Atmaca:
>> > Can you attach your PHP-FPM configuration file and NginX virtual
>> > host
>> > section for the site?
>>
>> Hello
>>
>> Debian stable (strech).
>>
>> packages:
>> php7.2-fpm
>> 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
>> server-side, HTML-embedded scripting language (FPM-CGI binary)
>> nginx 1.10.3-1+deb9u1
>> all small, powerful, scalable web/proxy server
>>
>
> This php7.2-fpm package is definitely not from the stable repository.
> php7.2 is not available in stretch and the latest php7.0-fpm version
> is 7.0.30-0+deb9u1
>
>> Fpm config:
>>
>> [www]
>> user = www-data
>> group = www-data
>> listen = /run/php/php7.2-fpm.sock
>> listen.owner = www-data
>> listen.group = www-data
>> pm = dynamic
>> pm.max_children = 5000
>> pm.start_servers = 4000
>> pm.min_spare_servers = 4000
>> pm.max_spare_servers = 5000
>
> Could it be that starting 4000 fpm processes eats up all your resources
> and starting the fpm-master process fails? Can you please check if the
> php7.2-fpm service is running?
>
> Can you try something like this?
>
> pm = dynamic
> pm.max_children = 250
> pm.start_servers = 25
> pm.min_spare_servers = 15
> pm.max_spare_servers = 30
>
>> -- Nginx config:
>>
>> server {
>> listen 80;
>>
>> root /home/website/;
>> index index.php index.html index.htm;
>> client_max_body_size 256M;
>> server_name y.br
>>
>>
>>
>> location = /favicon.ico {
>> log_not_found off;
>> access_log off;
>>}
>>
>>
>> if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
>> return 403;
>> }
>>
>>
>> location ~*
>> \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$ {
>> expires 365d;
>> add_header Cache-Control "public";
>>}
>>
>>
>> location ~* \.(?:css|js)$ {
>> expires 1y;
>> access_log off;
>>
Re: fpm problem
On 07/26/2018 03:37 PM, Gokan Atmaca wrote:
>> Can you attach your PHP-FPM configuration file and NginX virtual host
>> section for the site?
>
> Hello
>
> Debian stable (strech).
>
> packages:
> php7.2-fpm
> 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
> server-side, HTML-embedded scripting language (FPM-CGI binary)
> nginx 1.10.3-1+deb9u1
> all small, powerful, scalable web/proxy server
>
> Fpm config:
>
> [www]
> user = www-data
> group = www-data
> listen = /run/php/php7.2-fpm.sock
> listen.owner = www-data
> listen.group = www-data
> pm = dynamic
> pm.max_children = 5000
> pm.start_servers = 4000
> pm.min_spare_servers = 4000
> pm.max_spare_servers = 50> --- Nginx config:
The configuration is set for very high load. You've used Unix sockets.
My advice is :
- try to use TCP/IP for communication between FPM and NginX
Kind regards
Georgi
>
> server {
> listen 80;
>
> root /home/website/;
> index index.php index.html index.htm;
> client_max_body_size 256M;
> server_name y.br
>
>
>
> location = /favicon.ico {
> log_not_found off;
> access_log off;
>}
>
>
> if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
> return 403;
> }
>
>
> location ~* \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$ {
> expires 365d;
> add_header Cache-Control "public";
>}
>
>
> location ~* \.(?:css|js)$ {
> expires 1y;
> access_log off;
> add_header Cache-Control "public";
> }
>
>
> location / {
> try_files $uri $uri/ /index.php?q=$uri&$args;
>
>
> }
>
> location /my{
> allow x.x.x.x;
> allow x.x.x.x;
> if ($scheme = http) { return 301 https://$host$request_uri;}
> deny all;
> error_page 403 error403.html;
> try_files $uri $uri/ /index.php?q=$uri&$args;
> }
>
> location ~ \.php$ {
> try_files $uri =404;
> fastcgi_split_path_info ^(.+\.php)(/.+)$;
> fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
> fastcgi_index index.php;
> fastcgi_param SCRIPT_FILENAME $request_filename;
> include fastcgi_params;
> }
>
> gzip on;
> gzip_disable "msie6";
> gzip_comp_level 9;
> gzip_min_length 1100;
> gzip_buffers 16 8k;
> gzip_proxied any;
> gzip_types
> text/plain
> text/css
> text/js
> text/xml
> text/javascript
> application/javascript
> application/x-javascript
> application/json
> application/xml
> application/xml+rss;
>
> On Thu, Jul 26, 2018 at 2:56 PM, Georgi Naplatanov wrote:
>> On 07/26/2018 12:45 PM, Gokan Atmaca wrote:
What happened after you did what the warning message suggested to do?
>>>
>>> 502 gateway timeout
>>
>> Can you attach your PHP-FPM configuration file and NginX virtual host
>> section for the site?
>>
>> Which Debian branch are you using - stable, testing or unstable? Where
>> did you get PHP 7.2 packages from ?
>>
>> Kind regards
>> Georgi
>>
>>> On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber
>>> wrote:
Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST:
> we are receiving an error as below. what is the cause of the problem?
> (32g ram , 240g ssd , xeon 8 cor cpu)
>
> WARNING: [pool www] seems busy (you may need to increase
> pm.start_servers, or pm.min/max_spare_servers), spawning 32 children,
> there are 1 idle, and 143 total children
What happened after you did what the warning message suggested to do?
--
Regards
mk
>>>
>>
Re: fpm problem
Am Donnerstag, den 26.07.2018, 15:37 +0300 schrieb Gokan Atmaca:
> > Can you attach your PHP-FPM configuration file and NginX virtual
> > host
> > section for the site?
>
> Hello
>
> Debian stable (strech).
>
> packages:
> php7.2-fpm
> 7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
> server-side, HTML-embedded scripting language (FPM-CGI binary)
> nginx 1.10.3-1+deb9u1
> all small, powerful, scalable web/proxy server
>
This php7.2-fpm package is definitely not from the stable repository.
php7.2 is not available in stretch and the latest php7.0-fpm version
is 7.0.30-0+deb9u1
> Fpm config:
>
> [www]
> user = www-data
> group = www-data
> listen = /run/php/php7.2-fpm.sock
> listen.owner = www-data
> listen.group = www-data
> pm = dynamic
> pm.max_children = 5000
> pm.start_servers = 4000
> pm.min_spare_servers = 4000
> pm.max_spare_servers = 5000
Could it be that starting 4000 fpm processes eats up all your resources
and starting the fpm-master process fails? Can you please check if the
php7.2-fpm service is running?
Can you try something like this?
pm = dynamic
pm.max_children = 250
pm.start_servers = 25
pm.min_spare_servers = 15
pm.max_spare_servers = 30
> -- Nginx config:
>
> server {
> listen 80;
>
> root /home/website/;
> index index.php index.html index.htm;
> client_max_body_size 256M;
> server_name y.br
>
>
>
> location = /favicon.ico {
> log_not_found off;
> access_log off;
> }
>
>
> if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
> return 403;
> }
>
>
> location ~*
> \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$ {
> expires 365d;
> add_header Cache-Control "public";
> }
>
>
> location ~* \.(?:css|js)$ {
> expires 1y;
> access_log off;
> add_header Cache-Control "public";
> }
>
>
> location / {
> try_files $uri $uri/ /index.php?q=$uri&$args;
>
>
> }
>
> location /my{
> allow x.x.x.x;
> allow x.x.x.x;
> if ($scheme = http) { return 301 https://$host$request_uri;}
> deny all;
> error_page 403 error403.html;
> try_files $uri $uri/ /index.php?q=$uri&$args;
> }
>
> location ~ \.php$ {
> try_files $uri =404;
> fastcgi_split_path_info ^(.+\.php)(/.+)$;
> fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
> fastcgi_index index.php;
> fastcgi_param SCRIPT_FILENAME $request_filename;
> include fastcgi_params;
> }
>
> gzip on;
> gzip_disable "msie6";
> gzip_comp_level 9;
> gzip_min_length 1100;
> gzip_buffers 16 8k;
> gzip_proxied any;
> gzip_types
> text/plain
> text/css
> text/js
> text/xml
> text/javascript
> application/javascript
> application/x-javascript
> application/json
> application/xml
> application/xml+rss;
>
> On Thu, Jul 26, 2018 at 2:56 PM, Georgi Naplatanov
> wrote:
> > On 07/26/2018 12:45 PM, Gokan Atmaca wrote:
> > > > What happened after you did what the warning message suggested
> > > > to do?
> > >
> > > 502 gateway timeout
> >
> > Can you attach your PHP-FPM configuration file and NginX virtual
> > host
> > section for the site?
> >
> > Which Debian branch are you using - stable, testing or unstable?
> > Where
> > did you get PHP 7.2 packages from ?
> >
> > Kind regards
> > Georgi
> >
> > > On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber
> > > wrote:
> > > > Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST:
> > > >
> > > > > we are receiving an error as below. what is the cause of the
> > > > > problem?
> > > > > (32g ram , 240g ssd , xeon 8 cor cpu)
> > > > >
> > > > > WARNING: [pool www] seems busy (you may need to increase
> > > > > pm.start_servers, or pm.min/max_spare_servers), spawning 32
> > > > > children,
> > > > > there are 1 idle, and 143 total children
> > > >
> > > > What happened after you did what the warning message suggested
> > > > to do?
> > > >
> > > > --
> > > > Regards
> > > > mk
> > > >
>
>
--
Jens Sauer
smime.p7s
Description: S/MIME cryptographic signature
Re: fpm problem
> Can you attach your PHP-FPM configuration file and NginX virtual host
> section for the site?
Hello
Debian stable (strech).
packages:
php7.2-fpm
7.2.6-1+0~20180611145758.22+stretch~1.gbpe20e8b amd64
server-side, HTML-embedded scripting language (FPM-CGI binary)
nginx 1.10.3-1+deb9u1
all small, powerful, scalable web/proxy server
Fpm config:
[www]
user = www-data
group = www-data
listen = /run/php/php7.2-fpm.sock
listen.owner = www-data
listen.group = www-data
pm = dynamic
pm.max_children = 5000
pm.start_servers = 4000
pm.min_spare_servers = 4000
pm.max_spare_servers = 5000
--- Nginx config:
server {
listen 80;
root /home/website/;
index index.php index.html index.htm;
client_max_body_size 256M;
server_name y.br
location = /favicon.ico {
log_not_found off;
access_log off;
}
if ($http_user_agent ~* LWP::Simple|wget|libwww-perl) {
return 403;
}
location ~* \.(?:jpg|jpeg|gif|png|ico|cur|gz|svg|svgz|mp4|ogg|ogv|webm|htc)$ {
expires 365d;
add_header Cache-Control "public";
}
location ~* \.(?:css|js)$ {
expires 1y;
access_log off;
add_header Cache-Control "public";
}
location / {
try_files $uri $uri/ /index.php?q=$uri&$args;
}
location /my{
allow x.x.x.x;
allow x.x.x.x;
if ($scheme = http) { return 301 https://$host$request_uri;}
deny all;
error_page 403 error403.html;
try_files $uri $uri/ /index.php?q=$uri&$args;
}
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $request_filename;
include fastcgi_params;
}
gzip on;
gzip_disable "msie6";
gzip_comp_level 9;
gzip_min_length 1100;
gzip_buffers 16 8k;
gzip_proxied any;
gzip_types
text/plain
text/css
text/js
text/xml
text/javascript
application/javascript
application/x-javascript
application/json
application/xml
application/xml+rss;
On Thu, Jul 26, 2018 at 2:56 PM, Georgi Naplatanov wrote:
> On 07/26/2018 12:45 PM, Gokan Atmaca wrote:
>>> What happened after you did what the warning message suggested to do?
>>
>> 502 gateway timeout
>
> Can you attach your PHP-FPM configuration file and NginX virtual host
> section for the site?
>
> Which Debian branch are you using - stable, testing or unstable? Where
> did you get PHP 7.2 packages from ?
>
> Kind regards
> Georgi
>
>> On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber
>> wrote:
>>> Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST:
>>>
we are receiving an error as below. what is the cause of the problem?
(32g ram , 240g ssd , xeon 8 cor cpu)
WARNING: [pool www] seems busy (you may need to increase
pm.start_servers, or pm.min/max_spare_servers), spawning 32 children,
there are 1 idle, and 143 total children
>>>
>>> What happened after you did what the warning message suggested to do?
>>>
>>> --
>>> Regards
>>> mk
>>>
>>
>
Re: fpm problem
On 07/26/2018 12:45 PM, Gokan Atmaca wrote: >> What happened after you did what the warning message suggested to do? > > 502 gateway timeout Can you attach your PHP-FPM configuration file and NginX virtual host section for the site? Which Debian branch are you using - stable, testing or unstable? Where did you get PHP 7.2 packages from ? Kind regards Georgi > On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber > wrote: >> Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST: >> >>> we are receiving an error as below. what is the cause of the problem? >>> (32g ram , 240g ssd , xeon 8 cor cpu) >>> >>> WARNING: [pool www] seems busy (you may need to increase >>> pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, >>> there are 1 idle, and 143 total children >> >> What happened after you did what the warning message suggested to do? >> >> -- >> Regards >> mk >> >
Re: fpm problem
> What happened after you did what the warning message suggested to do? 502 gateway timeout On Thu, Jul 26, 2018 at 11:50 AM, Markus Schönhaber wrote: > Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST: > >> we are receiving an error as below. what is the cause of the problem? >> (32g ram , 240g ssd , xeon 8 cor cpu) >> >> WARNING: [pool www] seems busy (you may need to increase >> pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, >> there are 1 idle, and 143 total children > > What happened after you did what the warning message suggested to do? > > -- > Regards > mk >
Re: fpm problem
Gokan Atmaca, Do 26 Jul 2018 08:46:46 CEST: > we are receiving an error as below. what is the cause of the problem? > (32g ram , 240g ssd , xeon 8 cor cpu) > > WARNING: [pool www] seems busy (you may need to increase > pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, > there are 1 idle, and 143 total children What happened after you did what the warning message suggested to do? -- Regards mk
Re: fpm problem
> we are receiving an error as below. what is the cause of the problem? > (32g ram , 240g ssd , xeon 8 cor cpu) nginx 1.10 version and php7.2-fpm... On Thu, Jul 26, 2018 at 9:46 AM, Gokan Atmaca wrote: > Hello > > we are receiving an error as below. what is the cause of the problem? > (32g ram , 240g ssd , xeon 8 cor cpu) > > WARNING: [pool www] seems busy (you may need to increase > pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, > there are 1 idle, and 143 total children
Re: fpm problem
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Thu, Jul 26, 2018 at 09:46:46AM +0300, Gokan Atmaca wrote: > Hello > > we are receiving an error as below. what is the cause of the problem? > (32g ram , 240g ssd , xeon 8 cor cpu) > > WARNING: [pool www] seems busy (you may need to increase > pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, > there are 1 idle, and 143 total children No idea -- looks like some web server stuff. If you want to help your helpers, you might want to say *where* you found that "error" (it's actually a warning) (log file, what?), and perhaps what program issued it (web server? perhaps Apache? -- we can only guess...). Cheers - -- tomás -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.12 (GNU/Linux) iEYEARECAAYFAltZb5kACgkQBcgs9XrR2kbAzgCeMZrHiFvvFTN6U2Sjrk2/Hb9V fNYAnjZ9QBxaDoyiCS8eF/WFzPmUT+k9 =mOJM -END PGP SIGNATURE-
fpm problem
Hello we are receiving an error as below. what is the cause of the problem? (32g ram , 240g ssd , xeon 8 cor cpu) WARNING: [pool www] seems busy (you may need to increase pm.start_servers, or pm.min/max_spare_servers), spawning 32 children, there are 1 idle, and 143 total children
