Load balancing...almost working...

[keita_mog]

Hi everyone!

I have the following servers:

Load director 1:
hostname: reilbq03
eth0 - 10.94.233.22
eth0:0 - 10.94.233.109

Load director 2:
hostname: reilbq04
eth0 - 10.94.233.91
eth0:0 - 10.94.233.109

Real web-server 1:
hostname: reilbq01
eth0 - 10.94.233.36
lo:0 - 10.94.233.109

Real web-server 2:
hostname: reilbq02
eth0 - 10.94.233.24
lo:0 - 10.94.233.109


- I've already enabled ip forwarding on both directors

- I've already configured heartbeat nodes and all is running:

================================================== =
crm_mon -i 5

============
Last updated: Wed Jan 5 19:36:51 2011
Current DC: reilbq04 (xxx)
2 Nodes configured.
2 Resources configured.
============

Node: reilbq03 (xxx): online
Node: reilbq04 (xxx): online

Clone Set: stonith_clone
ssh_stonith:0 (stonith:external/ssh): Started reilbq03
ssh_stonith:1 (stonith:external/ssh): Started reilbq04
Resource Group: load_balancer
vip (ocf::heartbeat:IPaddr2): Started reilbq04
ldirector (ocf::heartbeat:ldirectord): Started reilbq04

================================================== ==



- I've already configured ldirectord and seems ok

================================================== ==
reilbq03:/usr/local/apache2/logs # ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 10.94.233.109:80 wlc
-> 10.94.233.24:80 Route 1 0 0
-> 10.94.233.36:80 Route 1 0 0
================================================== ==



- Even the weight's responding via the responses sent to the corresponding web servers.


BUT...when trying to open in the browser 10.94.233.109, I keep getting the

"•Connection to 10.94.233.109 Failed
The system returned:

(110) Connection timed out
"

Any recommendation?

[keita_mog]

As an update, I stopped ldirector on both directors

=====================
/etc/init.d/ldirectord stop
=====================

And restarted heartbeat (so the heartbeat would be the one to start and stop the service) on both directors:

=====================
/etc/init.d/heartbeat stop

/etc/init.d/heartbeat start
=====================


Now, when opening in a browser 10.94.233.109 (the virtual ip address), I got this error:

=====================
The system returned:

(113) No route to host
=====================

What I noticed is that in spite ipvsadm -Ln shows that the weight is 1, InActConn grows everytime I try to open the address in a browser

================================================== ====
reilbq03:/usr/local/apache2/bin # ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 10.94.233.109:80 wlc
-> 10.94.233.24:80 Route 1 0 23
-> 10.94.233.36:80 Route 1 0 23
================================================== ====