pqf in httpd

Move procmgr_init_spawn_cmd() out of a loop

Remove an unnecessary memset() call in procmgr_init_spawn_cmd()

remove an unnecessary check in a nest loop of ap_create_environment()

New protocol to avoid too much sleep, improve performance while stress testing

1. procmgr_send_spawn_cmd() now return a status code from PM, so process handler now know the spawn request is denyed or not.

2. if a new process is created, no sleep is needed.

3. if no process is created, sleep a while

PR: 53693

Submitted by: Ryan Pan <panqingfeng gmail>

Modify function names for a better understanding

1. procmgr_post_spawn_cmd -> procmgr_send_spawn_cmd

2. procmgr_peek_cmd -> procmgr_fetch_cmd

Handle exception while fcgid_create_privileged_process() fail. base on the patch from Igor Seletskiy

Bug fix, Bug 47873 - unreliable coordination between daemon and request thread for BusyTimeout processing

New protocol is:

1. Process node's diewhy init with FCGID_DIE_KILLSELF

2. Process manager set diewhy to FCGID_DIE_BUSY_TIMEOUT and gracefully kill process while busy timeout

3. Process manager forced kill process while busy timeout and diewhy is FCGID_DIE_BUSY_TIMEOUT

4. Request handler move process node to error list if node's diewhy is FCGID_DIE_BUSY_TIMEOUT

Add fcgid extension to mod_status

procnode->requests_handled does not increase in some cases
set wrong config name "FastCgiAuthenticator" and "FastCgiAuthenticatorAuthoritative"

while loading these configurations

apply a patch from Gabriel Barazer

Fix miner Sanity check bug

modify for win32 compile

use file to store large http request

The on the way up the spawn control counter increases with the process

starts and terminations, in other words with site load.

On the way down the spawn control counter is decreased by time elapsed.

With the up count based on usage and the down count based on time, a

difficult situation is presented for heavily loaded sites. The up count

will always go up faster than down count if you have terminations. This is

especially true for sites that use mod_fcgid for PHP and have more than 500

req/sec. The standard is to exit PHP after 500 requests, so if you are

serving more than 500 requests per second, you bleed up on the spawn count

without bounds (unless termination score is set to zero). This bleed up

means you will eventually hit your spawn limit and new spawns will be

limited by the passage of time and not the capacity of your server.

This patch allows the user to configure a scale factor for decreasing the

spawn counter with the passage of time. For sites with large capacity

this patch allows the administrator to scale down at a rate greater than

one per second.


Support apr_shm_remove() in httpd.2.0.X

Support configuration "PassHeader"

Support arguments in FCGIWrapper

Bug fix. Should not pass respond to auth checkers

Support arguments in FCGIWrapper

Check running user is root or not, while suexec is enabled

Check running user is root or not, while suexec is enabled

Support new config SharememPath, which specifies the location of share memory path

Support arguments in FCGIWrapper

Support arguments in FCGIWrapper

Support arguments in FCGIWrapper

Bug fix. Authoritative flag reversed

Change new child spawn log from notice to info

FCGIWrapper disallowed in .htaccess

Add missing config.m4 and for static linking

    • ?
    • ?