Re: failover without downtime?

On Thu 18 Feb 2016, Stig Brautaset wrote:
> Amir Dafny-Man <[email protected]> writes:
> >
> > I have no experience with haproxy but this may be solved via a configuration that replicates session information across haproxy instances.
> > Look at stick tables https://cbonte.github.io/haproxy-dconv/configuration-1.5.html#3.5
> 
> Thank you very much! This might indeed be very helpful for us. I assume
> that with that in place we should be able to just fail over keepalived
> by either stopping the keepalived master or bumping the priority on the
> current keepalived backup? The latter would be preferrable for us, as I
> assume it would affect only a particular VIP, rather than all the VIP
> managed by this keepalived instance.

I also have haproxy behind keepalived.

In my keepalived config I have:


vrrp_script chk_haproxy {
        script "/usr/bin/killall -0 haproxy"
        interval 10
}

vrrp_instance haproxy {
	[...]
	track_script {
		# add 20 points of prio if OK
                chk_haproxy weight 20
        }
}


This means that keepalived will stop being master if haproxy is not
running. You might want tot reduce the check interval though.


Paul

------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140