[Linux-cluster] lvs + nanny + piranha problem, server not working

[GS R]

On 11/6/07, jimmy.nimo at pranical.com <jimmy.nimo at pranical.com> wrote:
>
>
>
>
> Hello, thanks for reading my email,


Welcome :)

 I have a problem and perphaps someone in the
> list can work it out
>
> I have 3 working redhat 4 update 5 trying to work with piranha+lvs+nanny,
> 2 Real
> Server and 1 LVS Router, the router have 2 NIC, 1 with IP 172.16.247.130
> (eht0),
> and the other with 10.11.12.9(eth1), the real server are 10.11.12.11 and
> 10.11.12.12, I want to put the virtual server to run in the ip
> 172.16.247.150
> (eth0:1) and the gateway in the real servers are 10.11.12.10 (eth1:1 in
> the
> router server),


Since you have configured LVS- NAT your gateway on the real servers should
be 10.11.12.9

I create the lvs.cf with piranha and start pulse, but, NOTHING
> happens, piranha create the 10.11.12.10 and the 172.16.247.150 virtual
> interfaces (I can conect to the gui of piranha in
> http://172.16.247.150:3636)
> but can't do in the port 80. here are my lvs.cf
>
> serial_no = 137
> primary = 172.16.247.130
> primary_private = 10.11.12.9
> service = lvs
> backup_active = 0
> backup = 172.16.247.131
> backup_private = 10.11.12.11
> heartbeat = 1
> heartbeat_port = 539
> keepalive = 3
> deadtime = 10
> network = nat
> nat_router = 10.11.12.10 eth1:1
> nat_nmask = 255.255.255.0
> debug_level = NONE
> monitor_links = 0
> virtual http {
>     active = 1
>     address = 172.16.247.150 eth0:1
>     vip_nmask = 255.255.255.0
>     port = 80
>     use_regex = 0
>     load_monitor = ruptime


You should start 'rwhod' service on the real servers since you opted for
'load_monitor' as ruptime.

    scheduler = rr
>     protocol = tcp
>     timeout = 5
>     reentry = 5
>     quiesce_server = 0
>     server uno {
>         address = 10.11.12.11
>         active = 1
>         weight = 1
>     }
>     server dos {
>         address = 10.11.12.12
>         active = 1
>         weight = 1
>     }
> }


Rest configuration looks fine.

and I don't know why, but the ipsvadmin don't show the entrys of the real
> servers:
>
> [root at node1 ~]# ipvsadm -Ln
> IP Virtual Server version 1.2.0 (size=4096)
> Prot LocalAddress:Port Scheduler Flags
> -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
> TCP  172.16.247.150:80 rr


Make sure you have the 80 port up and running on your real servers.

can anyone help me? (sorry for my bad english)



-GSR
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/linux-cluster/attachments/20071106/963be5b7/attachment.htm>