Re: Director Failover

Hi Sebastien,

Does a connection *always* survive a director failover
?

I did a couple of tests by stopping the primary
director pulse and the backup director takes over.
Before doing that, I can see the telnet connection in
the backup director using ipvsadm -lc. But the telnet
connection got disconnected most of the times though I
did encounter some successful cases. So it seems to me

it's not *always* that a connection can survive a
director failover. Am I right ?

P.S. I had the syncdaemon up in both primary and
backup director as shown below:

root      7933     1  0 15:00 ?        00:00:00 [ipvs
syncbackup]

My lvs.cf as follow:

service = lvs
primary = 192.168.123.127
primary_private = 192.168.123.127
backup = 192.168.123.109
backup_private = 192.168.123.109
backup_active = 1
heartbeat = 1
heartbeat_port = 1050
syncdaemon = 1
keepalive = 6
deadtime = 18
reservation_conflict_action = preempt
debug_level = NONE
network = direct

virtual server1 {
         address = 192.168.123.110 eth0:1
         vip_nmask = 255.255.255.0
         active = 1
         timeout = 5
         reentry = 10
         port = 23
         send_program = "/etc/sysconfig/ha/monscript
%h"
         load_monitor =
"/etc/sysconfig/ha/ruptime_piranha.sh %h"
         expect = "OK"
         scheduler = rr
         pmask = 255.255.255.255
         protocol = tcp

         server main {
                 address = 192.168.123.109
                 active = 0
         }

         server email {
                 address = 192.168.123.198
                 active = 1
         }
}

Thanks
Chee Chye
 
 --- Sébastien Bonnet <[email protected]>
wrote: > > I'm testing Piranha+Direct routing with
telnet
> service using a
> > primary director and a backup director which is
> also acting as
> > a realserver (due to shortage of machines).
> 
> Then the stock piranha is missing a feature and
> you'll have to apply a
> patch (I've provided many times on this list)
> against piranha-0.7.0 and
> handle what you've foreseen in a later post
> (handling of VIP on eth/lo
> during failover).
> 
> > I telnet from another machine to the VIP. I then
> stop the pulse in
> > the primary director and the backup director take
> over. Would the
> > existing telnet connection be disconnected ?
> 
> If you've started the sync daemon (manually with
> your version of
> piranha, might be automatic with the patch), you
> telnet session *should*
> survive a director failover.
> 
> -- 
> Sébastien Bonnet
>   Centre de contacts - Experian France
> 
> 
> 
> _______________________________________________
> Piranha-list mailing list
> [email protected]
>
https://listman.redhat.com/mailman/listinfo/piranha-list 

__________________________________________________
Do You Yahoo!?
Promote your business from just $5 a month!
http://sg.biztools.yahoo.com