Most of connections are in 'READ' state?
Added by quanta over 13 years ago
Hi,
I have 3 servers behind the LVS. 2 of them have a lot of connections in 'READ' state:
551 connections
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhrrrrrrrrrr
h
fastcgi.active-requests: 1
fastcgi.backend.localhost.0.connected: 1432
fastcgi.backend.localhost.0.died: 0
fastcgi.backend.localhost.0.disabled: 0
fastcgi.backend.localhost.0.load: 0
fastcgi.backend.localhost.0.overloaded: 0
fastcgi.backend.localhost.1.connected: 8639
fastcgi.backend.localhost.1.died: 0
fastcgi.backend.localhost.1.disabled: 0
fastcgi.backend.localhost.1.load: 0
fastcgi.backend.localhost.1.overloaded: 0
fastcgi.backend.localhost.2.connected: 57906
fastcgi.backend.localhost.2.died: 0
fastcgi.backend.localhost.2.disabled: 0
fastcgi.backend.localhost.2.load: 0
fastcgi.backend.localhost.2.overloaded: 0
fastcgi.backend.localhost.3.connected: 356790
fastcgi.backend.localhost.3.died: 0
fastcgi.backend.localhost.3.disabled: 0
fastcgi.backend.localhost.3.load: 1
fastcgi.backend.localhost.3.overloaded: 0
fastcgi.backend.localhost.load: 1
fastcgi.requests: 424767
while the remaining server:
398 connections
kCCkkCCkkCCCkkCCCCCCCCkCCkkkkCkkkkCCCCCCCkCCCCkCkk
CkCCCCCCCCkCCkkCkkCkCkCkkCCCCCCCCCCCCCkCCCCkCCCkCC
CCCCkCCkCkkkCCkkCCkkCCkkkkkkkkCkkkkkkCCCCCCkCkkCkC
CCCkkCCkkCCCkkCCkkkkkCkCCCCCkCCkkkCkkCkCCkkkkkCCCC
kkCCCCCCCkkCCCkCkkCCCkCCCkkkCCkCCkCkCCCkkkCkkCCCCk
CCCCCCCkCCkCCCCkkCCCCCCCkkCCkCCCCkCkCkkkCkkkCkkCCC
CkCkCCCCCCCCkCCCkkCCkCCCCkCCCCCkkkkCkkCkCCkCCCkCCC
CkkCkCCCCCCCCCkCCCCCCCCkCkkCCCkCkkCCCCCCCkkkkkkh
fastcgi.active-requests: 0
fastcgi.backend.localhost.0.connected: 33538
fastcgi.backend.localhost.0.died: 0
fastcgi.backend.localhost.0.disabled: 0
fastcgi.backend.localhost.0.load: 0
fastcgi.backend.localhost.0.overloaded: 0
fastcgi.backend.localhost.1.connected: 158896
fastcgi.backend.localhost.1.died: 0
fastcgi.backend.localhost.1.disabled: 0
fastcgi.backend.localhost.1.load: 0
fastcgi.backend.localhost.1.overloaded: 0
fastcgi.backend.localhost.2.connecned: 688799
fastcgi.backend.localhost.2.died: 0
fastcgi.backend.localhost.2.disabled: 0
fastcgi.backend.localhost.2.load: 0
fastcgi.backend.localhost.2.overloaded: 0
fastcgi.backend.localhost.3.connected: 2346483
fastcgi.backend.localhost.3.died: 0
fastcgi.backend.localhost.3.disabled: 0
fastcgi.backend.localhost.3.load: 0
fastcgi.backend.localhost.3.overloaded: 0
fastcgi.backend.localhost.load: 0
fastcgi.requests: 3227716
I have asked the other team who manage the LVS, they said it works fine. Are there any problems with lighttpd on this server?
Replies (3)
RE: Most of connections are is 'READ' state? - Added by nitrox over 13 years ago
This is missing the needed config snippets and version of lighty and php. read-state most certainly means your php is f$%"&ed up.
RE: Most of connections are in 'READ' state? - Added by quanta over 13 years ago
Sorry, this is my php config section:
fastcgi.server = ( ".php" =>
( "localhost" =>
(
"socket" => "/tmp/php-fastcgi.socket",
"bin-path" => "/usr/bin/php-cgi",
"max-procs" => 4,
"bin-environment" => (
"PHP_FCGI_CHILDREN" => "128",
"PHP_FCGI_MAX_REQUESTS" => "10000"
),
"bin-copy-environment" => (
"PATH", "SHELL", "USER"
),
"broken-scriptfilename" => "enable"
)
)
)
lighttpd version 1.4.22
PHP 5.2.9 with XCache 1.2.2
The version of lighttpd and PHP on the server which has normal state is newer, 1.4.26 and 5.2.13 respectively (the config is the same).
RE: Most of connections are in 'READ' state? - Added by nitrox over 13 years ago
First, do you really need 512 php processes/children? Sounds like overkill if you aren´t Facebook. Use sth. like max-procs 2 and children 16 (max-procs should be close to 1 if you use it with an opcode cache like xcache). Second, there have been massive changes and bugfixes between 1.4.22 and 1.4.28 (current one). It might solve all your problems if you do sane settings on php and upgrade lighty.