On Aug 8, 2007, at 1:25 PM, Peter Rothman wrote:
I have the situation listed below. We had a h/w problem on one OSA
and we
lost connectivity to some of our Linux machines.
Is it possible to have Linux 'failover' within these parameters?
1. z/VM 5.2 and z/OS share 2 OSAs
2. The OSAs are connected to different 6509s.
3. I have 'part' of a subnet on each OSA which I have to use for
z/VM
and the Linux machines (SUSE Linux Enterprise Server 10)
4. Subnet on OSA 1 is 172.31.149.0/26 with gw of 172.31.149.2.
My 'part'
is 172.31.149.43 to .62
5. Subnet on OSA 2 is 172.31.149.64/26 with gw of 172.31.149.66. My
'part' is 172.31.149.107 to .126
6. Our 2 production Linux machines attach directly to the OSAs -
one on
each OSA.
Um, not within THOSE parameters.
But there are at least two ways to get failover here.
The easy one is to make the OSAs part of a VSWITCH configuration, and
give VM and the Linux guests each a single address in the VSWITCH,
and let VSWITCH handle the failover for you.
If you can't do that, then rather than giving each Linux machine one
OSA, give each an interface on EACH OSA, then define a VIPA address
for each Linux guest, use quagga to announce the VIPA via OSPF (like
you already do for VM), and then refer to each machine via its VIPA
address.
Adam
----------------------------------------------------------------------
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
