Re: Testing lvs using Piranha 0.7.0-3 in Linux 8.0
Sébastien Bonnet <[email protected]>
| Newsgroups | gmane.linux.redhat.piranha |
|---|---|
| Message-ID | <[email protected]> |
> Thanks Peter. Can I use Linux 7.3 to generate the
> lvs.cf and run piranha on Linux 8.0 ?
Of course you can. But I think Peter was saying that 7.3 is way more
stable than 8.0 and you should use 7.3 instead of 8.0, whatever you're
doing.
> I got the following message when starting pulse:
>
> [root@main ha]# /etc/rc.d/init.d/pulse start
> Starting pulse:
> [FAILED]
Have a look at /var/log/messages and you'll certainly get clues to
what's you problem. If not start again pulse -v -c /etc/...
> Below is my lvs.cf file (most of the options have been
> commented).
As far as I remember, you can't comment out lines. You have to remove
then. Also, many of the lines you've commented out are mandatory. Below
is what I'd expect to be minimal :
> [root@main ha]# cat lvs.cf
service = lvs
primary = 192.168.123.109
backup = 0.0.0.0
backup_active = 0
heartbeat = 0
heartbeat_port = 1050
keepalive = 6
deadtime = 18
reservation_conflict_action = preempt
debug_level = NONE
rsh_command = rsh
network = direct
virtual server1 {
address = 192.168.123.110 eth0:1
vip_nmask = 255.255.255.0
active = 1
load_monitor = none
timeout = 5
reentry = 10
port = 23
scheduler = rr
persistent = 60
pmask = 255.255.255.255
protocol = tcp
server Real1 {
address = 192.168.123.198
active = 1
weight = 1
}
}
vip_nmask as been corrected from 255.255.255.255 to 255.255.255.0
load_monitor has been set to none, but I remember I fixed a problem with
that. You might try with uptime, eventhough it won't after the overall
system.
As for the send and expect string, I can't recall if nanny will just try
to connect and then disconnect if you don't specify anything. You might
try to specify the empty string for both.
/var/log/messages will be your best friend to understand what's going
on. Don't forget to start pulse with -v.
--
Sébastien Bonnet
Centre de contacts - Experian France