Hi,
telegraf inputs phpfpm doesn't handle '?' char.
In my nginx setup the pool name is a variable like:
# PHP-FPM Monitoring section
set $pool "default";
if ($arg_pool) {
set $pool $arg_pool;
}
location ~ "/fpm-(status|ping)" {
include /etc/nginx/fastcgi_params;
fastcgi_pass unix:/var/run/php-fpm_$pool.sock;
fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;
....
}
with this config you can call the pool name directly with:
http://localhost/fpm-status?pool=<mysuperpoolname>
Relevant telegraf.conf:
put in your config to use phpfpm inputs like :
[phpfpm]
urls = ["http://localhost/fpm-status?pool=pool0","http://localhost/fpm-status?pool=pool1"]
System info:
telegraf 1.0.1
Expected behavior:
127.0.0.1 - - [16/Oct/2016:22:39:00 +0200] "GET /fpm-status?pool=pool0 HTTP/1.1" 200 215 "-" "Go-http-client/1.1"
127.0.0.1 - - [16/Oct/2016:22:39:00 +0200] "GET /fpm-status?pool=pool1 HTTP/1.1" 200 206 "-" "Go-http-client/1.1"
Actual behavior:
127.0.0.1 - - [16/Oct/2016:22:26:00 +0200] "GET /fpm-status HTTP/1.1" 502 166 "-" "Go-http-client/1.1"
127.0.0.1 - - [16/Oct/2016:22:26:00 +0200] "GET /fpm-status HTTP/1.1" 502 166 "-" "Go-http-client/1.1"
Workaround
I replace '?' with %3F like:
[phpfpm]
urls = ["http://localhost/fpm-status%3Fpool=pool0","http://localhost/fpm-status%3Fpool=pool1"]
Regards,
Yann
Hi,
telegraf inputs phpfpm doesn't handle '?' char.
In my nginx setup the pool name is a variable like:
with this config you can call the pool name directly with:
Relevant telegraf.conf:
put in your config to use phpfpm inputs like :
System info:
telegraf 1.0.1
Expected behavior:
Actual behavior:
Workaround
I replace '?' with
%3Flike:Regards,
Yann