Home Page‎ > ‎

HeartBeat

To set up Heartbeat/Clustering two nodes are mandatory. You will also need 2 "Public IPs" and 2 "Internal IPs"



There are 3 configuration files that will need to be created and configured.

  • authkeys (encrypted passphrase, This will be identical on both master and slave)
  • haresources (defines master, This will be identical on both master and slave)
  • ha.cf ( defines master,slave,internal IP,etc, This will be different on each server)


1. Get updates

  • sudo aptitude update
  • sudo aptitude safe-upgrade

2. Configure Private IP

  • vi /etc/network/interface
  • add eth1 with private ip info
  • /etc/init.d/networking restart

2. Install Heartbeat

  • sudo aptitude install heartbeat

3. Navigate to Heartbeat Directory and create "authkeys"

  • vi /etc/heartbeat/authkeys
    auth 1
    1 sha1 yourpassphrasegoeshere
  • chmod 600 /etc/heartbeat/authkeys (must not be 'world' readable or else Heartbeat will fail)

4. Navigate to the Heartbeat Directory and create "haresources"

  • vi /etc/heartbeat/haresources
    masterhostname 172.16.110.XX/24 (public IP)

5. Navigate to Heartbeat Directory and create "ha.cf" on the master and edit the following lines

logfacility daemon
keepalive 2
deadtime 15
warntime 5
initdead 120
udpport 694
ucast eth1 ##This is where your Slave Private IP goes##
auto_failback on
node master ##Hostname of Master##
node slave ##Hostname of Slave##
respawn hacluster /usr/lib/heartbeat/ipfail
use_logd yes

6. Restart heartbeat

  • /etc/init.d/heartbeat restart

7. Navigate to Heartbeat Directory and create "ha.cf" on the slave and edit the following lines

logfacility daemon
keepalive 2
deadtime 15
warntime 5
initdead 120
udpport 694
ucast eth1 ##This is where your Master Private IP goes##
auto_failback on
node master ##Hostname of Master##
node slave ##Hostname of Slave##
respawn hacluster /usr/lib/heartbeat/ipfail
use_logd yes

8. Restart heartbeat

  • /etc/init.d/heartbeat restart

9. On Slave Server ping Master Public IP

  • ping -c2 172.16.110.XX

10. Shutdown Master Server ( Slave should take over by this time)

  • Try pinging Master again, ping -c2 172.16.110.XX ( You should receive reply)
Comments