Got a bit of an odd issue and I'm hoping someone can help me figure this out.
The set up is not ideal but here's the basic flow: Request goes to VIP1 on LVS on Blade A. It then routes it to one of 2 load balanced pairs on Blade A or B (works perfectly). That request then goes to one of 2 load balanced JBOSS instances which also reside on Blade A or B. So A--->AorB --->AorB. Weird I know, but the customer wanted it this way. So what I've found is that when a request is made from a Blade to one of the JBOSS instances on the same Blade, it hangs. If the request is made to the JBOSS instance on the other Blade it goes through ok. This can be tested with a simple wget. If I sit on Blade A and issue a wget to one of the backend servers on the same blade, it hangs. If I issue it to one of the backend servers not on the same blade, it goes through. I can't help but think there is something I have to set on the LVS side or even on the backend servers but I'm not sure what. Any ideas? _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
