LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: my LVS cluster

To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: my LVS cluster
From: Erik Elmshauser <erike@xxxxxxxx>
Date: Mon, 9 Oct 2000 12:34:28 -0500 (CDT)
On Mon, 9 Oct 2000, Joseph Mack wrote:

> If the 2nd webserver was being LVS'ed you wouldn't be able to get to it   
> from the director. Since you have no entry in ipvsadm for port 81, the
> director won't LVS it.
> 
> > Is there anything really obviously wrong with this setup?
> 
> Since you got it to work on port 80, you should be able to reproduce the
> setup again on port 81. I would assume that you've made some simple
> mistake.
> 

ok.  here is an edited version of the config I am using.  I have removed
comments and altered the valid IPs and moved the FTP vserver out from
between the web vserver sections, but made no other changes.

-----------------------------------
# SECTION 1 - GLOBAL SETTINGS
service = lvs
primary = 192.168.1.1
backup_active = 0
heartbeat = 0
heartbeat_port = 1050
rsh_command = ssh
network = nat
nat_router = 10.0.0.254 eth1

# SECTION 2 - VIRTUAL AND REAL SERVERS

virtual www {
        address = 192.168.1.1 eth0
        port = 80
        active = 1
        persistent = 60
        send = "GET / HTTP/1.0\r\n\r\n"
        expect = "HTTP"
        load_monitor = uptime
        reentry = 10
        server cluster1 {
                address = 10.0.0.1
                active = 1
        }
        server cluster2 {
                address = 10.0.0.2
                active = 1
        }
        server cluster3 {
                address = 10.0.0.3
                active = 1
        }
}

virtual web {
        address = 192.168.1.1 eth0
        port = 81
        active = 1
        persistent = 60
        send = "GET / HTTP/1.0\r\n\r\n"
        expect = "HTTP"
        load_monitor = uptime
        server cluster1 {
                address = 10.0.0.1
                active = 1
        }
}

virtual ftp {
        address = 192.168.1.1 eth0
        port = 21
        active = 1
        send = "\r\n"
        expect = "220"
        send = "quit"
        load_monitor = uptime
        server cluster1 {
                address = 10.0.0.1
                active = 1
        }
}
-----------------------------------

when I start pulse on the director (I have no backup) then I get 4 nannys
and ipvsadm produces the following output:

# /usr/sbin/ipvsadm 
IP Virtual Server version 0.9.16 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port          Forward Weight ActiveConn InActConn
TCP  balance:ftp wlc
  -> cluster1:ftp                Masq    1      0          0         
TCP  balance:www wlc persistent 60
  -> cluster3:www                Masq    1      0          0         
  -> cluster2:www                Masq    1      0          0         
  -> cluster1:www                Masq    1      0          0         


I am able to ftp and get web pages through LVS with this
configuration.  and when I am working from the director I can point a
browser to cluster1:81 and get a page back, so apache is going
strong.  anything here make it obvious what my mistake is, or any ideas
what I should try?

        --Erik



<Prev in Thread] Current Thread [Next in Thread>