ipfail fifos, ipfail require 'nice_failback on'

[Wallwork, Nathan]

On Wed, 28 Aug 2002, Kevin Dwyer wrote:
> If you mean adding so-called 'ping nodes', it's fairly simple.  You add a
> line to your ha.cf that says something like:  ping 1.1.1.1 2.2.2.2 3.3.3.3
> etc.
> 
> However, heartbeat won't failover anything simply because a ping node
> died.  I have written a plugin that does this which may be helpful to you,
> if I understand the logic of your network correctly.  Info here:
> http://pheared.net/~kevin/ipfail/

Great, I've now got ipfail running, and doing most everything I want.

I had to create the ipfail.req and ipfail.rsp fifos, as described in 
your inital announcement at 
	http://marc.theaimsgroup.com/?l=linux-ha&m=102087295903692&w=2
but I didn't find any details on how to do this.  Here's what I did, 
for anyone else with the same question:

        cd /usr/var/lib/heartbeat/api
        mkfifo ipfail.req
        mkfifo ipfail.rsp
        chgrp haclient ipfail.req ipfail.rsp
        chmod 600 ipfail.req ipfail.rsp

Also, ipfail apparently requires that nice_failback be enabled in ha.cf
with a line like this:

	nice_failback on

otherwise ha-log gets a message saying
	heartbeat ... info: Standby mode only implemented when nice_failback on