On Wednesday, 11/02/2005 at 09:12 CST, Adam Thornton <[EMAIL PROTECTED]> wrote:
> Oh, I'd say that chatter is still necessary. > > But what you do is define two different TCPIP VSWITCH controllers, on > (by definition) different physical OSA ports, with different cables > connecting them to the switch(es, ideally). Eh? What "by definition"? Define two controllers and don't worry about it any more. Just make sure that you have automation watching for Controller Death and restart it ASAP. The two controllers will provide the needed controller redundancy for any and all VSWITCHes you create. Details in the z/VM 5.1 Getting Started With Linux book. > Then configure two eth interfaces on each Linux guest, one coupled to > each VSWITCH. Then run quagga to advertise a dummy interface > spanning those two via OSPF. That way you can lose one OSA card, one > VSWITCH controller, or one VSWITCH interface on any guest without any > interruption of service. You can even pull the same trick with the > VM TCPIP interface itself, by giving your TCPIP (not necessarily a > VSWITCH controller) machine two virtual NICs, each coupled to a > different VSWITCH. And a reminder that if you have backup OSAs on a VSWITCH to be sure to plug them into different [physical] switches trunked together. Then if you lose a switch, the VSWITCH can keep running. (And plug them into different power circuits, too.) Alan Altmark z/VM Development IBM Endicott ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
