On Monday, 04/06/2009 at 04:15 EDT, Shimon Lebowitz <[email protected]> wrote: > > ===LAN==+== 10.0/16 ========== > |OSA > +-------------+ > | VSWITCH | > +-------------+ > |VNIC > +----------------------------+ +-----------------+ > | 10.0.7.153 | | VMunderVM | > | VM 5.3 Test | | +-------------+ | > | 10.1.99.53|--CTC----+ VMTCPIP | | > +----------------------------+ | | 10.1.31.253 | | > | +-------------+ | > > +-----------------+ > > 1) The network guy had prepared LAN routing for 10.1.31.x to 10.0.7.153 > 2) the vswitch shows PRIROUTER when I do a Q VSWITCH > 3) a workstation on the LAN can ping 10.0.7.153 > 4) when logged in on 10.0.7.153 I can ping the workstation, and 10.1.31.253 > 5) BUT... the workstation only gets timeouts when pinging 10.1.31.253. > > So, this is basically a test of whether I can route to another network via > VM on a vswitch, and it did not work. > Obviously, I cannot throw away the VM4 system until I can get the VM5 > to give connectivity to our VSE. (Yes, when we upgrade VSE next, it will also > connect directly to the Vswitch, with a new IP, and no more routing!) > > I am at a loss as to who is breaking the chain here. How can I know > if the routie set up to 10.0.7.153 is working? Or that the OSA > is really sending the request to the vswitch? Or is VM not forwarding via > the CTC to 2nd level TCPIP? > Any suggestions as to what to do/check next are welcome. :-)
You didn't post any NETSTAT/IFCONFIG output, so it's hard to say. Did you check your TCPIP consoles for errors? Miguel can confirm, but I think z/VM 5.2 tightened the rules on p2p subnet masks. (No more 255.255.255.255) Anyway: 1. Check NETSTAT GATE on both TCP/IPs. 2. Turn on MORETRACE for IPUP, IPDOWN, and ICMP on the routing system 3. Do the same for the endpoint system. Alan Altmark z/VM Development IBM Endicott
