Monday, August 22, 2016

CentOS and RHEL 7: Install Linux, Apache MPM, MariaDB, PHP (LAMP) Stack

CentOS and RHEL 7: Install Linux, Apache MPM, MariaDB, PHP (LAMP) Stack

Check CentOS version:

# cat /etc/redhat-release

CentOS Linux release 7.1.1503 (Core)

Set up Network:

# ip addr

2: eno16777736: <broadcast> mtu 1500 qdisc pfifo_fast state UP qlen 1000

# vi /etc/sysconfig/network-scripts/ifcfg-eno16777736

BOOTPROTO=none

IPADDR=192.168.6.9
PREFIX=24
GATEWAY=192.168.6.1
DNS1=8.8.8.8
DNS2=8.8.4.4

PEERDNS=yes
ONBOOT=yes
USERCTL=no

Note: USERCTL=no // Non-root users are not allowed to control this device.

# systemctl restart network

# ip addr

# /bin/ipcalc --netmask 192.168.6.9/24

NETMASK=255.255.255.0

# cat /etc/resolv.conf

Change hostname:

# hostnamectl status

# hostnamectl set-hostname cent-dev.local

# hostnamectl status

# cat /etc/hostname

If you are using Amazon's AWS EC2 instance, append the following string at the bottom of the file to ensure that the hostname is preserved between restarts/reboots:

# vim /etc/cloud/cloud.cfg

preserve_hostname: true

More info: https://aws.amazon.com/premiumsupport/knowledge-center/linux-static-hostname-rhel7-centos7/

Set up timezone:

# timedatectl
# timedatectl list-timezones
# timedatectl set-timezone America/Vancouver
# timedatectl

Set up date and time:

# date +%Y%m%d -s "20081128"
# date +%T -s "10:13:13"

To sync date and time automatically:

# yum -y update

# yum -y install ntp

# ntpdate 0.us.pool.ntp.org

Note: ntpdate is deprecated as of September 2012

You can change which ntp server to use:

# vi /etc/ntp.conf

Set ntpd to start up on boot time:

# systemctl enable ntpd.service
# systemctl restart ntpd.service

# ps auxww|grep -i ntpd
# ntpstat
# timedatectl

Set the Hardware Clock to the current System Time:

# hwclock --systohc

Note: hwclock is a utility for accessing the hardware clock. Hardware clock is independent of the operation system you use and works even when the machine is shut down. This program is used to find out the time from the hardware clock and set the system time at boot time.

Update the ~/.bashrc configuration:

# vi ~/.bashrc

### alias
alias rm='rm -i'
alias cp='cp -i'
alias mv='mv -i'

alias gs='git status'
alias gc='git commit -a -m "up"'
alias gp='git push'
alias gpp='git pull'

alias ls='ls --color=auto'
alias ll='ls -la'
alias h='history'

### ls with color (try "ls --color=auto").
#export CLICOLOR=1 # Use colors (if possible)
#export LSCOLORS="ExGxFxdxCxDxDxBxBxExEx"

### set up a clean UTF-8 environment
### run: locale command
export LC_ALL=en_US.UTF-8
export LANG=en_US.UTF-8
export LANGUAGE=en_US.UTF-8

### display history command with date and time
export HISTTIMEFORMAT="%m/%d/%y %T "

### Prompt
PS1='\[\e[0;32m\]\u@\h \w \$\[\e[0m\] '

#######
# Note: on Ubuntu, xterm-256color may be in different place, try this:
# find /lib/terminfo /usr/share/terminfo -name "*256*"
# Note: tmux respects screen-256color
#######
if [ -e /usr/share/terminfo/x/xterm-256color ]; then
  export TERM='xterm-256color'
else
  export TERM='xterm-color'
fi

### Make bash check its window size after a process completes
shopt -s checkwinsize

# Source global definitions
if [ -f /etc/bashrc ]; then
  . /etc/bashrc
fi

# source ~/.bashrc

Update the ~/.inputrc configuration:

# vi ~/.inputrc

### enable filename tab auto-completion
set show-all-if-ambiguous on
set show-all-if-unmodified on

### if you don't want case-sensitivity
#set completion-ignore-case on

### bash history completion to complete what's already on the line
### arrow up
"\e[A": history-search-backward
### arrow down
"\e[B": history-search-forward

Make sure there is a swap space on your system:

# cat /proc/meminfo | grep -i swap

SwapCached:           36 kB
SwapTotal:       2097148 kB
SwapFree:        2095160 kB

Note: if your system does not have the swap space, please refer to http://blog.ijun.org/2015/04/add-swap-to-amazon-ec2-instance-ebs.html for more information.

Install EPEL and additional repositories on CentOS and Red Hat:

# yum -y install wget
# yum -y install epel-release

http://blog.ijun.org/2014/11/install-epel-and-additional.html
https://fedoraproject.org/wiki/EPEL

Subscribing to the IUS Community Project Repository:

# curl 'https://setup.ius.io/' -o setup-ius.sh
# bash setup-ius.sh

# ls -l /etc/yum.repos.d/ius*

-rw-r--r--. 1 root root 1150 Apr 16  2015 /etc/yum.repos.d/ius-archive.repo
-rw-r--r--. 1 root root 1131 Apr 16  2015 /etc/yum.repos.d/ius-dev.repo
-rw-r--r--. 1 root root 1073 Apr 16  2015 /etc/yum.repos.d/ius.repo
-rw-r--r--. 1 root root 1150 Apr 16  2015 /etc/yum.repos.d/ius-testing.repo

To find which package provides the ifconfig command:

# yum provides ifconfig

To get the ifconfig command into our system:

# yum -y install net-tools

# ifconfig | awk '/inet /{print $2}'

Install tmux:
# yum -y install tmux

# vi ~/.tmux.conf

# tmux
or
# tmux a -d

Install vim-enhanced:

# vim --version
-syntax
-python

# yum -y install vim-enhanced

# vim --version
+syntax
+python

# yum list installed | grep -i vim
# yum info vim-enhanced

Install tree:

# yum -y install tree

Install fortune game:

# yum -y install fortune-mod.x86_64

Install Glances:

# yum -y install glances

Install Git:

# yum -y install git

Install NetCat:

# yum -y install nmap-ncat
# nc localhost 8080

Install firewalld:

# yum -y install firewalld

(optional)
# yum install firewall-config

Note: If you don't mind using a GUI you could use firewall-config instead. If you need something for the console you will have to use firewall-cmd instead.

# systemctl enable firewalld

# systemctl restart firewalld

Install MariaDB:

# yum install mariadb-server

Setting UTF8 defaults for MySQL:

We recommend against MySQL's utf8 character set, since it does not support 4-byte unicode characters, and strings containing them will be truncated. This is fixed by the newer utf8mb4 character set.

# vim /etc/my.cnf

[mysqld]
# Version 5.5.3 introduced "utf8mb4", which is recommended
collation-server = utf8mb4_general_ci # Replaces utf8_general_ci
character-set-server = utf8mb4 # Replaces utf8

default-storage-engine = InnoDB

max_allowed_packet = 16M

# This option makes InnoDB to store each created table into its own .ibd file.
innodb_file_per_table

# Don't resolve hostnames. All hostnames are IP's or 'localhost'.
skip-name-resolve

#The number of simultaneous clients allowed.
max_connections = 200

# uncomment to disable the InnoDB storage engine
#skip-innodb

innodb_buffer_pool_size = 4G
innodb_additional_mem_pool_size = 16M
innodb_log_file_size = 256M
innodb_log_buffer_size = 8M

# systemctl enable mariadb.service

# systemctl restart mariadb.service

# systemctl is-active mariadb.service

# /usr/bin/mysql_secure_installation

# mysql -u root -p

mysql> GRANT ALL PRIVILEGES ON *.* TO 'test'@'192.168.0.%' IDENTIFIED BY '123456';
mysql> FLUSH PRIVILEGES;

mysql> CREATE DATABASE mydb DEFAULT CHARACTER SET utf8 DEFAULT COLLATE utf8_general_ci;

Set up .my.cnf configuration file:

# touch ~/.my.cnf
# chmod 600 ~/.my.cnf
# vim ~/.my.cnf

[client]
host = localhost
port = 3306
user = root
password = MyPassword

Add New Rule to firewalld to allow access to MySQL:

# firewall-cmd --permanent --zone=trusted --add-source=192.168.0.1/32
# firewall-cmd --permanent --zone=trusted --add-port=3306/tcp
# firewall-cmd --reload
# firewall-cmd --zone=trusted --list-all

or

# systemctl restart firewalld.service

Check out the zone file to inspect the XML configuration:

# cat /etc/firewalld/zones/public.xml
# cat /etc/firewalld/zones/trusted.xml

Install Apache:

# yum install httpd mod_ssl

# systemctl status httpd

● httpd.service - The Apache HTTP Server
   Loaded: loaded (/usr/lib/systemd/system/httpd.service; disabled; vendor preset: disabled)

Note: it will tell you if the service starts on boot.

# systemctl enable httpd.service

# systemctl restart httpd.service

# systemctl reload httpd.service

# systemctl -l status httpd.service

# journalctl -xn

# systemctl is-active httpd.service

# apachectl configtest

# httpd -V

# apachectl graceful

httpd service default configuration files:

  • Default config file: /etc/httpd/conf/httpd.conf
  • Configuration files which load modules : /etc/httpd/conf.modules.d/ directory (e.g. PHP)
  • Select MPMs (Processing Model) as loadable modules [worker, prefork (default)] and event: /etc/httpd/conf.modules.d/00-mpm.conf
  • Default ports: 80 and 443 (SSL)
  • Default log files: /var/log/httpd/{access_log,error_log}

Set up a symbolic link:

# cd / ; ln -s var/www/html www

Install PHP:

# yum install php php-mysqlnd php-fpm php-gd php-mbstring php-pdo php-xml php-soap
# yum install php-pear php-devel pcre-devel gcc gcc-c++ make

# systemctl restart httpd.service

Note: you need php-devel pcre-devel gcc make for PHP APC. Try to add httpd-devel if failed.

You need the php-mcrypt to run Magento:

# yum install php-mcrypt

http://blog.ijun.org/2014/11/how-to-install-php-mcrypt-on-centos-7.html

Open port 80 firewall access:

# firewall-cmd --zone=public --add-port=http/tcp
# firewall-cmd --zone=public --add-port=https/tcp

# firewall-cmd --permanent --zone=public --add-port=http/tcp
# firewall-cmd --permanent --zone=public --add-port=https/tcp

# firewall-cmd --reload
or
# systemctl restart firewalld.service

# firewall-cmd --zone=public --list-all

Allow a IP address with a specific port:

# firewall-cmd --zone=public --add-rich-rule='rule family="ipv4" source address="1.2.3.4" port protocol="tcp" port="3306" accept'

To remove the rule:

# firewall-cmd --zone=public --remove-rich-rule='rule family="ipv4" source address="1.2.3.4" port protocol="tcp" port="3306" accept'

Block a IP address:

# firewall-cmd --zone=public --add-rich-rule='rule family="ipv4" source address="1.2.3.4" reject'

To remove the rule:

# firewall-cmd --zone=public --remove-rich-rule='rule family="ipv4" source address="1.2.3.4" reject'

If you get 403 forbidden error, then you probably have problem with SELinux, to deal with Security-Enhanced Linux (SELinux):

# namei -l /var/www/html/magento19

# ls -dZ /var/www/html
drwxr-xr-x. root root system_u:object_r:httpd_sys_content_t:s0 /var/www/html

# chcon -R --reference=/var/www/html /var/www/html/magento19
Or
# chcon -R --type=httpd_sys_content_t /var/www/html/magento19
Or for read and write permission:
# chcon -R -t httpd_sys_rw_content_t /var/www/html/magento19/app/etc

# ps auxwwZ | grep httpd
# ls -dZ /var/www/html/magento19/app/etc
# tail /var/log/audit/audit.log
# tail /var/log/messages

To turn off Security-Enhanced Linux (SELinux):

# setenforce 0

To turn on Security-Enhanced Linux (SELinux):

# setenforce 1

To get the status of a system running SELinux:

# sestatus

Set up Apache MPM and PHP-FPM:

With Apache 2.4, the official module to use is mod_proxy_fcgi instead of the ancient mod_fastcgi. That module, as well as mod_fcgid, were third party modules.

Note: mod_proxy_fcgi now supports network sockets since Apache 2.4.9 ( Unix socket support for mod_proxy_fcgi )

List built-in or shared modules:

# httpd -M | grep -iE 'proxy_module|proxy_fcgi_module|rewrite_module|mpm_event_module|deflate_module|vhost_alias_module|ssl_module'

 deflate_module (shared)
 rewrite_module (shared)
 vhost_alias_module (shared)
 mpm_event_module (shared)
 proxy_module (shared)
 proxy_fcgi_module (shared)
 ssl_module (shared)

List loaded modules:

# httpd -t -D DUMP_MODULES | grep -iE 'proxy_module|proxy_fcgi_module|rewrite_module|mpm_event_module|deflate_module|vhost_alias_module|ssl_module'

 deflate_module (shared)
 rewrite_module (shared)
 vhost_alias_module (shared)
 mpm_event_module (shared)
 proxy_module (shared)
 proxy_fcgi_module (shared)
 ssl_module (shared)

Edit mpm.conf:

# vim /etc/httpd/conf.modules.d/00-mpm.conf

Comment out the following line:

LoadModule mpm_prefork_module modules/mod_mpm_prefork.so

Uncomment the following line:

LoadModule mpm_event_module modules/mod_mpm_event.so

Make sure the following two lines exist:

# grep -E 'mod_proxy.so|mod_proxy_fcgi' /etc/httpd/conf.modules.d/00-proxy.conf

LoadModule proxy_module modules/mod_proxy.so
LoadModule proxy_fcgi_module modules/mod_proxy_fcgi.so

Add the "if checking" surround the following three lines:

# vim /etc/httpd/conf.d/php.conf

<IfModule mod_php5.c>
  <FilesMatch \.php$>
    SetHandler application/x-httpd-php
  </FilesMatch>
</IfModule>

Add the "if checking" surround the following two lines:

# vim /etc/httpd/conf.d/php.conf

<IfModule mod_php5.c>
  php_value session.save_handler "files"
  php_value session.save_path    "/var/lib/php/session"
</IfModule>

Change the following line:

# vim /etc/httpd/conf/httpd.conf

<Directory "/var/www/html">
  AllowOverride All

  # New directive needed in Apache 2.4.3: 
  Require all granted
</Directory>

Create and Edit the vhosts file:

# vim /etc/httpd/conf.d/httpd-vhosts.conf

#
# Use name-based virtual hosting.
#
NameVirtualHost *:80

#
# VirtualHost example:
# Almost any Apache directive may go into a VirtualHost container.
# The first VirtualHost section is used for all requests that do not
# match a ServerName or ServerAlias in any <VirtualHost> block.
#
<VirtualHost *:80>
    ServerAdmin webmaster@dummy-host.example.com
    DocumentRoot "/var/www/html/host_not_found"
    ErrorLog "/var/log/httpd/host_not_found-error_log"
    CustomLog "/var/log/httpd/host_not_found-access_log" common
</VirtualHost>

<VirtualHost *:80>
    ServerAdmin webmaster@dummy-host.example.com
    DocumentRoot "/var/www/html/magento1.9.1"
    ServerName us.centos.local

    ### Disable PHP script execution for this directory. We don't want to reverse-proxy this subdirectory.
    <Location "/var">
      ProxyPass !
    </Location>

    ### Disable PHP script execution for this directory. We don't want to reverse-proxy this subdirectory.
    <Location "/media">
      ProxyPass !
    </Location>

    ProxyPassMatch ^/(.*\.php(/.*)?)$ fcgi://127.0.0.1:9000/var/www/html/magento1.9.1/$1

    SetEnv MAGE_RUN_CODE "us_centos_local"
    SetEnv MAGE_RUN_TYPE "website"

    ### a request for / will need to be mapped to a resource on the fcgi backend. Failure to address this may cause a blank response, commonly known as a WSOD (White Screen of Death), especially if only a request URI containing the php extension is proxied, such as this example. The processing chain will first map a request for / to /index.php, then proxy to the PHP-FPM backend correctly.
    DirectoryIndex /index.php index.php index.html index.htm

    ErrorLog "/var/log/httpd/us.centos.local-error_log"
    CustomLog "/var/log/httpd/us.centos.local-access_log" common
</VirtualHost>

Setting up an SSL secured Web server:

<VirtualHost *:443>
    SSLEngine on
    SSLCertificateFile /etc/pki/tls/certs/mydomain.com.crt
    SSLCertificateKeyFile /etc/pki/tls/private/mydomain.com.key
    SSLCertificateChainFile /etc/pki/tls/certs/gd_bundle-mydomain.com.crt

    ServerAdmin webmaster@dummy-host.example.com
    DocumentRoot "/var/www/html/magento1.9.1"
    ServerName us.centos.local

    ### we don't want to reverse-proxy this subdirectory.
    <Location "/var">
      ProxyPass !
    </Location>

    ### we don't want to reverse-proxy this subdirectory.
    <Location "/media">
      ProxyPass !
    </Location>

    ProxyPassMatch ^/(.*\.php(/.*)?)$ fcgi://127.0.0.1:9000/var/www/html/magento1.9.1/$1

    SetEnv MAGE_RUN_CODE "us_centos_local"
    SetEnv MAGE_RUN_TYPE "website"

    ErrorLog "/var/log/httpd/us.centos.local443-error_log"
    CustomLog "/var/log/httpd/us.centos.local443-access_log" common
</VirtualHost>

Make sure the security context of the certification files are cert_t:

# cd /etc/pki/tls/certs/mydomain.com
# chcon -t cert_t *

If you do not see the correct security context, you will see the following message when restarting Apache:

Mar 16 23:58:00 ip-172-31-21-55 httpd[22680]: AH00526: Syntax error on line 19 of /etc/httpd/conf.d/httpd-vhosts.conf:
Mar 16 23:58:00 ip-172-31-21-55 httpd[22680]: SSLCertificateFile: file '/etc/pki/tls/certs/mydomain.com/b1210d10x4d812c5.crt' does not exist or is empty
Mar 16 23:58:00 ip-172-31-21-55 systemd[1]: httpd.service: main process exited, code=exited, status=1/FAILURE
Mar 16 23:58:00 ip-172-31-21-55 systemd[1]: Failed to start The Apache HTTP Server.
Mar 16 23:58:00 ip-172-31-21-55 systemd[1]: Unit httpd.service entered failed state.
Mar 16 23:58:00 ip-172-31-21-55 systemd[1]: httpd.service failed.

Restart Apache:

# systemctl reload httpd.service

# httpd -t -D DUMP_VHOSTS

To enable gzip compression:

# egrep 'deflate|header' /etc/httpd/conf.modules.d/00-base.conf

LoadModule deflate_module modules/mod_deflate.so
LoadModule headers_module modules/mod_headers.so

# httpd -t -D DUMP_MODULES | grep deflate

deflate_module (shared)

# vim /etc/httpd/conf.d/mod_deflate.conf

<IfModule mod_deflate.c>
  AddOutputFilterByType DEFLATE text/html text/plain text/xml text/css text/js application/javascript application/x-javascript application/x-httpd-php
</IfModule>

# systemctl restart httpd.service

Install Xdebug for debugging PHP:

# pecl install Xdebug

# vim /etc/php.d/xdebug.ini

zend_extension="/usr/lib64/php/modules/xdebug.so"

; When this setting is set to on, the tracing of function calls will be enabled just before the script is run. This makes it possible to trace code in the auto_prepend_file.
xdebug.auto_trace = 1
xdebug.trace_output_dir = "/tmp"
xdebug.collect_params = 4

; Enables Xdebug's profiler which creates files in the profile output directory. Those files can be read by KCacheGrind to visualize your data.
xdebug.profiler_enable = 1
xdebug.profiler_output_dir = "/tmp"

; Controls the protection mechanism for infinite recursion protection. The value of this setting is the maximum level of nested functions that are allowed before the script will be aborted.
xdebug.max_nesting_level = 100

; shows a human readable / computer readable trace file.
xdebug.trace_format = 0

; This setting tells Xdebug to gather information about which variables are used in a certain scope. This analysis can be quite slow as Xdebug has to reverse engineer PHP's opcode arrays. This setting will not record which values the different variables have, for that use xdebug.collect_params. This setting needs to be enabled only if you wish to use xdebug_get_declared_vars().
xdebug.collect_vars = 0

; When set to '1' the trace files will be appended to, instead of being overwritten in subsequent requests.
; Note: this option can be useful if you could not find your function calls anywhere.
xdebug.trace_options = 1

To search all other php modules:

# yum search php

Edit php.ini:

# vim /etc/php.ini

cgi.fix_pathinfo = 0

The default value is 1, which is an extremely insecure setting because it tells PHP to attempt to execute the closest file it can find if a PHP file does not match exactly. This basically would allow users to craft PHP requests in a way that would allow them to execute scripts that they shouldn't be allowed to execute.

Note: if I set it to "cgi.fix_pathinfo = 0", I would get "Access denied (403)" (see security.limit_extensions) or no input file specified error when setting up Magento. You can either:

1. commented out the cgi.fix_pathinfo = 0 line.
2. set cgi.fix_pathinfo = 1
3. try to set "security.limit_extensions = " in the /etc/php-fpm.d/www.conf file.

http://stackoverflow.com/questions/23390531/access-denied-403-for-php-files-with-nginx-php-fpm
http://serverfault.com/questions/627903/is-the-php-option-cgi-fix-pathinfo-really-dangerous-with-nginx-php-fpm

display_errors = On
log_errors = On
error_log = /var/log/php_errors.log

Note: make sure you do:

# touch /var/log/php_errors.log
# chmod 660 /var/log/php_errors.log
# chown root:apache /var/log/php_errors.log

# chcon -t httpd_log_t /var/log/php_errors.log
Or
# chcon -u system_u -t httpd_log_t /var/log/php_errors.log

Install XCache:

XCache is a fast, stable ​PHP opcode cacher that has been proven and is now running on production servers under high load. It is tested (on linux) and supported on all of the latest ​PHP release branches such as PHP_5_1 PHP_5_2 PHP_5_3 PHP_5_4 PHP_5_5. It is more stable than APC.

Warning: APC would cause segfault segmentation fault. Use XCache instead.

# yum install php-xcache xcache-admin

# systemctl restart httpd
# systemctl restart php-fpm
# php -v

PHP 5.4.16 (cli) (built: Oct 31 2014 12:59:36)
Copyright (c) 1997-2013 The PHP Group
Zend Engine v2.4.0, Copyright (c) 1998-2013 Zend Technologies
    with XCache v3.1.1, Copyright (c) 2005-2014, by mOo
    with XCache Optimizer v3.1.1, Copyright (c) 2005-2014, by mOo
    with XCache Cacher v3.1.1, Copyright (c) 2005-2014, by mOo
    with XCache Coverager v3.1.1, Copyright (c) 2005-2014, by mOo

# echo -n "Your Password" | md5sum

5afd8756ebeda48acf7eb645503dae60

# vim /etc/php.d/xcache.ini

[xcache]

xcache.admin.user = "admin name"
xcache.admin.pass = "5afd8756ebeda48acf7eb645503dae60"

xcache.size = 60M

; This number divides total cache size into threads in order to increase the efficiency. If you have 128M xcache.size and you set the count as 4, that means each thread will manage 32M size of cache.
xcache.count = 1

xcache.cacher = On
xcache.stat = On
xcache.optimizer = On

# cp -r /usr/share/xcache/ /var/www/html/

# systemctl restart php-fpm

http://localhost/xcache/

Install APC:

Warning: APC would cause segfault segmentation fault. Use XCache instead.

# pecl install apc

# vim /etc/php.d/apc.ini

; Enable APC for PHP
extension=apc.so
apc.enabled=1

; The number of seconds a cache entry is allowed to idle in a slot before APC dumps the cache
apc.ttl=72000
apc.user_ttl=72000
apc.gc_ttl=3600

; Size of memory for apc ( 1024 M)
apc.shm_size=1024M

; Enable apc stats.
apc.stat=1

; Enable APC for command line php operations.
apc.enable_cli=1

; Allow 2 seconds after a file is created before it is cached. This will prevent premature PHP pages to get cached.
apc.file_update_protection=2

; Maximum size of single file that apc can store.
apc.max_file_size=1M

; Maximum number of files APC can store ( rotation).
apc.num_files_hint=200000

; Maximum number of users data entries that APC can store.
apc.user_entries_hint=20000

Copy the apc.php file:

# cp /usr/share/pear/apc.php /var/www/html

Set up php-fpm:

Apach 2.4.8 mod_proxy: Added support for unix domain sockets as the backend server endpoint.

# vim /etc/php-fpm.d/www.conf

If you are using Apach 2.4.8 or above, please change the following line from:

listen = 127.0.0.1:9000

To:

listen = /var/run/php-fpm/php-fpm.sock

Now, there are different ways to actually forward requests for .php files to this module, ranging from everything (using ProxyPass) to very specific or rewritten files or patterns (using mod_rewrite with the [P] flag).

The method I chose (using ProxyPassMatch) lies somewhere in between these in complexity and flexibility, since it allows you to set one rule for all PHP content of a specific vhost, but will only proxy .php files (or URLs that contain the text .php somewhere in the request).

TCP socket (IP and port) approach

Edit the configuration for a vhost of your choice, and add the following line to it:

# vim /etc/httpd/conf.d/httpd-vhosts.conf

ProxyPassMatch ^/(.*\.php(/.*)?)$ fcgi://127.0.0.1:9000/path/to/your/documentroot/$1

### a request for / will need to be mapped to a resource on the fcgi backend. Failure to address this may cause a blank response, commonly known as a WSOD (White Screen of Death), especially if only a request URI containing the php extension is proxied, such as this example. The processing chain will first map a request for / to /index.php, then proxy to the PHP-FPM backend correctly.
    DirectoryIndex /index.php index.php index.html index.htm

Note: please do change /path/to/your/documentroot to for example /var/www/html/drupal8

Look confusing ? Let's run through it:

ProxyPassMatch

only proxy content that matches the specified regex pattern; in this case:

^/(.*\.php(/.*)?)$

from the documentroot onwards, match everything ending in .php (with the dot escaped), optionally followed by a slash and any continued path you like (some applications use this so-called PathInfo to pass arguments to the php script.)

The ^ (caret) and $ (dollar) signs are used to anchor both the absolute start and end of the URL, to make sure no characters from the request escape our pattern match.

The nested parentheses enable us to refer to the entire request-URI (minus the leading slash) as $1, while still keeping the trailing pathinfo optional.

fcgi://127.0.0.1:9000

forward via mod_proxy_fcgi, using the fastCGI protocol, to the port our php-fpm daemon is listening on.

This determines which fastcgi pool will serve requests proxied by this rule.

/path/to/your/documentroot/

IMPORTANT! This must exactly match the real filesystem location of your php files, because that is where the php-fpm daemon will look for them.

php-fpm just interprets the php files passed to it; it is not a web server, nor does it understand your web servers' namespace, virtualhost layout, or aliases.

IMPORTANT! Read the above again

$1

expands to the entire request-URI from the original request, minus the leading slash (because we already added that above.)

DirectoryIndex /index.php index.php index.html index.htm

Note: a request for / will need to be mapped to a resource on the fcgi backend. Failure to address this may cause a blank response, commonly known as a WSOD (White Screen of Death), especially if only a request URI containing the php extension is proxied, such as this example. The processing chain will first map a request for / to /index.php, then proxy to the PHP-FPM backend correctly.

unix domain socket (UDS) approach

Edit the configuration for a vhost of your choice, and add the following line to it:

# vim /etc/httpd/conf.d/httpd-vhosts.conf

ProxyPassMatch ^/(.*\.php(/.*)?)$ unix:/path/to/socket.sock|fcgi://127.0.0.1:9000/path/to/your/documentroot/

unix:/path/to/socket.sock

the path to your fpm socket

Note that with this approach, the captured request URI ($1) is not passed after the path

Enable php-fpm to start on boot:

# systemctl enable php-fpm.service

Start our PHP processor:

# systemctl restart php-fpm.service
# systemctl restart httpd.service

Test PHP:

# php -r "echo 'hi';";

Check the PHP-FPM setting:

<?php
echo phpinfo();
?>

You should see the message: Server API: FPM/FastCGI

# httpd -V

Server MPM:     event

To create a locked user account:

# useradd dev -m -c 'git user'

Unlock the account by issuing the passwd command to assign a password and set password aging guidelines:

# passwd dev

Add a new group called web:

# groupadd web

Add the dev user to the web group:

# usermod -a -G web dev

Show The Groups a User Is In:

# groups dev
# id -Gn dev

Find out the primary group of a user:

# getent group dev

To allow Apache to connect network and sendmail send email:

# setsebool -P httpd_can_network_connect 1
# setsebool -P httpd_can_sendmail 1

Note: -P in the above command means Persistent (across reboots)

# getsebool -a | grep -i httpd_can

httpd_can_network_connect --> on
httpd_can_sendmail --> on

# sestatus -b | grep httpd_can

httpd_can_network_connect                   on
httpd_can_sendmail                          on

Install Memcached:

# yum install memcached
# yum install php-pecl-memcached

# systemctl enable memcached.service
# systemctl restart memcached.service

# memcached-tool localhost:11211 display
# memcached-tool localhost:11211 stats
# memcached-tool localhost:11211 dump

Install redis server:

# yum install redis
# yum install php-pecl-redis

Two important redis server configuration files:

# less /etc/redis.conf
# less /etc/redis-sentinel.conf

Start the Redis server:

# systemctl start redis.service

Check the running status of Redis server:

# systemctl status redis.service

To test the installation of Redis:

# redis-cli ping

PONG

To enable Redis server at system's booting time:

# systemctl enable redis.service

To get the listening port 6379 of Redis server:

# ss -nlp | grep redis

To install Redis PHP extension:

# pecl install redis

Add the following line to /etc/php.d/redis.ini:

# echo 'extension=redis.so' >> /etc/php.d/redis.ini

Check to see if Redis PHP extension is installed:

# pecl list | grep redis

redis   2.2.7   stable

Restart Apache and PHP-FPM:

# systemctl restart httpd.service
# systemctl restart php-fpm.service

To see if Redis extension is being loaded by PHP:

# php -m | grep redis

To allow Apache to connect to the Redis server:

# setsebool -P httpd_can_network_connect 1

Note: If you have turned on Security-Enhanced Linux (SELinux), httpd scripts by default are not allowed to connect out to the network.

To list all Redis Databases:

# redis-cli info keyspace

To clear remove delete all data from a particular Redis database:

# redis-cli

127.0.0.1:6379> info keyspace
127.0.0.1:6379> select 0
127.0.0.1:6379> keys *
127.0.0.1:6379> flushdb
127.0.0.1:6379> keys *

To clear remove delete all data from all Redis database:

# redis-cli flushall

Dumping all key/value pairs in a Redis db:

# redis-cli -n 0 keys \*

# redis-cli -n 0 keys \* | xargs -n 1 redis-cli dump

Note: the 0 is the database number.

To store PHP sessions in Redis:

Storing PHP session files in RAM can be much more efficient than storing on disk and can also save some IO. To configure this, you should modify the main php.ini file and change session.save_handler to redis.

# vim /etc/php.ini

session.save_handler = redis

session.save_path = "tcp://127.0.0.1:6379"

# systemctl restart php-fpm

# php -r 'echo phpinfo();' | grep redis

Registered save handlers => files user redis
session.save_handler => redis => redis

# vim test.php

<?php
session_start();

$_SESSION['favcolor'] = 'green';

echo '<pre>' . print_r($_SESSION, TRUE) . '</pre>';
?>

# php test.php

# redis-cli info keyspace
# Keyspace
db0:keys=68,expires=39,avg_ttl=2977110
db2:keys=427,expires=427,avg_ttl=1856886

# redis-cli -n 0 keys \*| grep -i session

PHPREDIS_SESSION:vhauaf8qpdj146kirsbivrh4i7

Note: the 0 is the database number.

Redis setup hints

  • We suggest deploying Redis using the Linux operating system. Redis is also tested heavily on osx, and tested from time to time on FreeBSD and OpenBSD systems. However Linux is where we do all the major stress testing, and where most production deployments are working.
  • Make sure to set the Linux kernel overcommit memory setting to 1. Add vm.overcommit_memory = 1 to /etc/sysctl.conf and then reboot or run the command sysctl vm.overcommit_memory=1 for this to take effect immediately.
  • Make sure to disable Linux kernel feature transparent huge pages, it will affect greatly both memory usage and latency in a negative way. This is accomplished with the following command: echo never > sys/kernel/mm/transparent_hugepage/enabled.
  • Make sure to setup some swap in your system (we suggest as much as swap as memory). If Linux does not have swap and your Redis instance accidentally consumes too much memory, either Redis will crash for out of memory or the Linux kernel OOM killer will kill the Redis process.
  • Set an explicit maxmemory option limit in your instance in order to make sure that the instance will report errors instead of failing when the system memory limit is near to be reached.
  • If you are using Redis in a very write-heavy application, while saving an RDB file on disk or rewriting the AOF log Redis may use up to 2 times the memory normally used. The additional memory used is proportional to the number of memory pages modified by writes during the saving process, so it is often proportional to the number of keys (or aggregate types items) touched during this time. Make sure to size your memory accordingly.
  • Use daemonize no when run under daemontools.
  • Even if you have persistence disabled, Redis will need to perform RDB saves if you use replication, unless you use the new diskless replication feature, which is currently experimental.
  • If you are using replication, make sure that either your master has persistence enabled, or that it does not automatically restarts on crashes: slaves will try to be an exact copy of the master, so if a master restarts with an empty data set, slaves will be wiped as well.

Running Redis on EC2

  • Use HVM based instances, not PV based instances.
  • Don't use old instances families, for example: use m3.medium with HVM instead of m1.medium with PV.
  • The use of Redis persistence with EC2 EBS volumes needs to be handled with care since sometimes EBS volumes have high latency characteristics.
  • You may want to try the new diskless replication (currently experimetnal) if you have issues when slaves are synchronizing with the master.

Reference:

http://blog.ijun.org/2014_04_01_archive.html
http://blog.ijun.org/2014/11/configuring-magento-to-use-redis.html
http://blog.ijun.org/2014/12/install-apache-24-php-56-and-mysql-56.html
http://redis.io/topics/admin
http://www.cyberciti.biz/faq/howto-install-linux-apache-mariadb-php-lamp-stack-on-centos7-rhel7/
http://serverfault.com/questions/629937/centos-7-apache2-httpd-mod-fastcgi-installation-impossible
http://blog.famillecollet.com/post/2014/08/01/Apache-httpd-server-2.4.10-and-PHP-FPM-5.6-in-Fedora-21
https://wiki.apache.org/httpd/PHP-FPM
http://sharadchhetri.com/2014/10/04/install-redis-server-centos-7-rhel-7/
https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/6/html/Deployment_Guide/s1-networkscripts-interfaces.html
http://technovergence-en.blogspot.ca/2012/03/mysql-from-utf8-to-utf8mb4.html

No comments: