HBEATD(7)                                                            HBEATD(7)



NAME
       hbeatd -- a simple and lightweight heartbeat sensor and pulse program

SYNOPSIS
       hbeatd [ options... ]
       hbeatd [ -pv ] [ -d ip ] [ -P port ] [ -i interval ] [ -g group ]
       hbeatd [ -sv ] [ -P port ] [ -t tolerance ]

DESCRIPTION
       hbeatd  is  a  simple and lightweight program for sending and receiving
       heartbeats that can be used to check the status of nodes in  a  cluster
       of servers. The pulse program on the nodes sends udp packages to chosen
       sensor. The sensor collects the ip's along with the  server  group  and
       builds a list. Everytime a new heartbeat is received it will update the
       timestamp for the node and check the list for dead nodes (nodes with  a
       timestamp  older than the tolerance level). If a node is dead, a script
       located at /etc/hbeatd/rc.d will run.

OPTIONS
       Unless specifically stated otherwise, options  are  applicable  in  all
       operating modes.

       -p     Run hbeatd in pulse mode.

       -s     Run hbeatd in sensor mode.

       -v     Run in verbose mode, not as daemon.

       -d ip   Destination  ip  address  (only  available  in pulse mode).  The
              default is 127.0.0.1.

       -P port Destination port number (only available  in  pulse  mode).   The
              default is 6220.

       -i interval
              The  interval  in milliseconds between pulses (only available in
              pulse mode).  The default is 1000 (1 second).

       -g group
              A group name ex. 'web' or 'db',  good  when  building  lists  of
              available nodes for deployment and maintenance etc. (only avail-
              able in pulse mode).  The default is 'misc'.

       -t tolerance
              The number of seconds before reporting that  the  node  is  dead
              (only available in sensor mode).  The default is 4.

EXAMPLES
       hbeat -d 192.168.0.50
              Start hbeatd in pulse mode.

       hbeatd -p -i 800 -v
              Start  hbeatd  in  pulse  mode  without  daemonizing it. Sending
              heartbeats with 800 milliseconds interval.

       hbeatd -s
              Start hbeatd in sensor mode.

       hbeatd -s -t 20
              Start hbeatd in sensor mode with a tolereance of 20 seconds.

       hbeatd -s -P 8910 -v
              Start hbeatd in sensor mode on port 8910 without daemonizing  it
              (good for debugging).

FILES
       /etc/hbeatd/rc.d
              An executable file that runs everytime hbeatd founds a new, dead
              or resurrected node.  The script or program can take three argu-
              ments; status, ip and group.
              Example sh-script:

              if [ "$1" = "dead" ]; then
                   echo "dead: $2($3)" >> /tmp/hbeatd.log
              elif [ "$1" = "up" ]; then
                   echo "up: $2($3)" >> /tmp/hbeatd.log
              elif [ "$1" = "new" ]; then
                   echo "new: $2($3)" >> /tmp/hbeatd.log
              fi

AUTHOR
       Jack Engqvist Johansson <jack.johansson@comfirm.se>



Hbeatd Version 1.3.1           14 September 2012                     HBEATD(7)