README.xxxhasrv

README file for {start|stop|monitor}hasrv scripts.
You will have to modify each of these scripts,
 to start and stop your desired services.


starthasrv:
	Starts the services you want, on an active node.
	Be paranoid, and check to see that they properly start.
	Return 0 on good start, non-zero (eg: 1) on fail


stophasrv:
	The yin to starthasrv's yang.
	Stops the services on a node.
	You MUST stop services before you exit from the script.
	"Failure is NOT an option".
	Return value of the script is not checked.
	Do not exit the script until the services are stopped
	This script is called both when a clean stop is requested, AND
	after monitorhasrv detects an error. 

monitorhasrv:
	Monitor everything you are running, with this script.
	Either everything is good, or it is time to take things down and
	fail over to another node.
	Return 0, and everything stays happy.
	Return non-0, and the demon will call stophasrv, and mark
	this node as errored.
	At that point, you have to kill and restart the HA demon to bring
	things back on-line.
	This is a safety feature. In theory, it is better to avoid corrupted
	starts, than to bounce back and forth in and endless loop between the
	nodes vainly trying to restart services. 

alerthasrv:
	By default, does nothing. This gets called whenever freehad changes
	state. For example, when a standby node goes active, its state
	changes from "STANDBY" to "RUNNING"