Hey all,

Looking for a little advice for something that I'm hoping should be relatively simple.

Right now, we're running a load balancing server in front of a few Windows web servers. The Linux load balancer is running ldirectord, and is configured with a handful of virtual IP addresses on a single interface. The Windows systems are configured with their own IP addresses, and also have loopback devices set up with the same addresses that the load balancer has. This set up has been working great.

Now, I'm trying to do a similar set up using Linux systems as the web server, but I'm not having any luck finding a good writeup on how to do this. Setting up virtual adapters (ifcfg-lo:0) does not seem to work. I've seen recommendations on installing arptables_jf, setting the arp_announce and arp_ignore options in sysctl, but haven't been able to get them to work. Tomorrow I plan on trying to install a physical second network card in a real web server and set the shared IP on that interface to see if I can get something to work.

I guess what I'm looking for in a simple writeup on getting this working using only ldirectord with the real web servers running Linux (CentOS 5 for the LB, CentOS6 for the web servers). Any suggestions?

Thanks!

"Does not seem to work" is not the technical kind of assessment of a problem anyone here can do anything with. Could be more efficient instead to run us through the steps and post relevant diagnostics / (error) message output?..