High Availability HA LAMP server

Home

Tips & Resources

Written by Travis Nuske

Last Updated on Wednesday, 15 September 2010

If you feel this tip has saved you time or effort, please consider buying us a cuppa coffee to keep things going!

Install 2 Fedora 13 systems as normal

update and upgrade both systems

Configure networking on both nodes

System -> Preferences -> Network Connections

webnode1 10.71.128.71

webnode2 10.71.128.72

Edit /etc/hosts

127.0.0.1   localhost
10.71.128.71   webhost1 webhost1.mps.sch
10.71.128.72   webhost2 webhost2.mps.sch

Update both nodes

System -> Administration -> Update Manager

Remove SELinux

Install NTP

install with synaptic

edit /etc/ntp.conf

change server to your ntp server

Install Distributed Block Device drbd and Heartbeat

drbd8-utils and heartbeat with synaptic

Set permissions on drbd files

chgrp haclient /sbin/drbdsetup
chmod o-x /sbin/drbdsetup
chmod u+s /sbin/drbdsetup
chgrp haclient /sbin/drbdmeta
chmod o-x /sbin/drbdmeta
chmod u+s /sbin/drbdmeta

fdisk -l to see name of second disk, in my case:

   root@webnode2:~# fdisk -l



   Disk /dev/sda: 21.5 GB, 21474836480 bytes

   255 heads, 63 sectors/track, 2610 cylinders

   Units = cylinders of 16065 * 512 = 8225280 bytes

   Sector size (logical/physical): 512 bytes / 512 bytes

   I/O size (minimum/optimal): 512 bytes / 512 bytes

   Disk identifier: 0x0002c932



       Device Boot      Start         End      Blocks   Id System

   /dev/sda1   *           1        2497    20051968   83 Linux

   /dev/sda2            2497        2611      916481    5 Extended

   /dev/sda5            2497        2611      916480   82 Linux swap / Solaris



   Disk /dev/sdb: 53.7 GB, 53687091200 bytes

   255 heads, 63 sectors/track, 6527 cylinders

   Units = cylinders of 16065 * 512 = 8225280 bytes

   Sector size (logical/physical): 512 bytes / 512 bytes

   I/O size (minimum/optimal): 512 bytes / 512 bytes

   Disk identifier: 0x00000000



   Disk /dev/sdb doesn't contain a valid partition table

resource lamp {
        protocol C;

        handlers {
        pri-on-incon-degr "echo o > /proc/sysrq-trigger ; halt -f";
        pri-lost-after-sb "echo o > /proc/sysrq-trigger ; halt -f";
        local-io-error "echo o > /proc/sysrq-trigger ; halt -f";
        outdate-peer "/usr/lib/heartbeat/drbd-peer-outdater -t 5";
        }

        startup {
        degr-wfc-timeout 120;
        }

        disk {
        on-io-error detach;
        }

        net {
        cram-hmac-alg sha1;
        shared-secret "password";
        after-sb-0pri disconnect;
        after-sb-1pri disconnect;
        after-sb-2pri disconnect;
        rr-conflict disconnect;
        }

        syncer {
        rate 100M;
        verify-alg sha1;
        al-extents 257;
        }

        on webnode1 {
        device /dev/drbd0;
        disk    /dev/sdb;
        address 10.71.128.71:7788;
        meta-disk internal;
        }

        on webnode2 {
        device /dev/drbd0;
        disk    /dev/sdb;
        address 10.71.128.72:7788;
        meta-disk internal;
        }
}

Initialse meta data on disks

[node1]drbdadm create-md lamp
[node2]drbdadm create-md lamp

Start drbd on both servers

service drbd start

setup webnode1 as the primary and start the sync

[node1]drbdadm -- --overwrite-data-of-peer primary lamp

you can watch the sync progress with the cat command:

watch cat /proc/drbd

Format and Mount new disk

[node1] mkfs.ext4 /dev/drbd0
[node1] mkdir -p /srv/data
[node1] mount /dev/drbd0 /srv/data

Insatall lamp files on both servers

I used synaptic selecting apache2, mysql-server php5 and php5-mysql

Remove Apache2 startup scripts

[node1]update-rc.d -f apache2 remove

[node2]update-rc.d -f apache2 remove

Remove DRBD starup scripts

update-rc.d -f drbd remove on both nodes

Edit startup script to prevent mysql from running at boot

Open /etc/init/mysql.conf

change the start on section so that it reads: start on runlevel [!0123456]

This will prevent mysql from running at boot, but allow the service mysql start command to still work properly.

Reboot each server

Run mysql manually to create data

/etc/ini.d/mysql start

On node 1 set drbd as primary

drbdadm primary lamp

and mount on /srv/data

mount /dev/drbd0 /srv/data

# Create location to store files
[node1]mkdir -p /srv/data/etc
[node1]mkdir -p /srv/data/var/lib
[node1]mkdir -p /srv/data/var/log

Move Apache2 and php5 files

# Move files to new location
[node1]mv /etc/apache2 /srv/data/etc
[node1]mv /etc/php5 /srv/data/etc
[node1]mv /var/lib/php5 /srv/data/var/lib
[node1]mv /var/www /srv/data/var
[node1]mv /var/log/apache2 /srv/data/var/log

# Link to new location
[node1]ln -s /srv/data/etc/apache2 /etc/apache2
[node1]ln -s /srv/data/etc/php5 /etc/php5
[node1]ln -s /srv/data/var/lib/php5 /var/lib/php5
[node1]ln -s /srv/data/var/www /var/www
[node1]ln -s /srv/data/var/log/apache2 /var/log/apache2

At this point you could reboot the server, make node1 primary, mount /dev/drdb0 and start apache2

etc/ha.d/ha.cf

logfacility local0
keepalive 2
deadtime 30
warntime 10
initdead 45
bcast eth0
node webnode1
node webnode2
auto_failback on
respawn hacluster /usr/lib/heartbeat/ipfail
logfile /var/log/heartbeat.log