[lvs-users] LVS does not route/forward after a few days

[Shaun Chong]

Hi,

I'm running LVS-DR (Direct Routing) on CentOS 5.3
I'm still very new to LVS.

After setting everything up, things seem to run fine:

IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP virtualip:80 wrr
-> realserver1:80 Route 1 11 134
-> realserver2:80 Route 1 12 139
-> realserver3:80 Route 1 11 143
-> realserver4:80 Route 1 14 145
-> realserver5:80 Route 1 15 140

However, after a few days of running, it doesn't seem to route/forward
anymore:

IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP virtualip:80 wrr
-> realserver1:80 Route 1 0 0
-> realserver2:80 Route 1 0 0
-> realserver3:80 Route 1 0 0
-> realserver4:80 Route 1 0 0
-> realserver5:80 Route 1 0 0

Checking my Apache access_log, I see that realserver4 is handling all the
requests, and the other servers are idling. 
ipvsadm -L shows 0 ActiveConn and 0 InActConn for realserver4 and the other
servers... which is weird; since realserver4 seems to be doing some work.
And why does the load balancer stop routing after a few days? Whenever this
happens, I execute a "service pulse restart" and it seems to work fine for
the next few days again.

Is there any way at all to diagnose this?

Regards,

Shaun Chong