Massimiliano Belardi wrote:
Guys,
I've a question. I've performed several installation of Quagga with VIPA
using two DIFFERENT subnet for the real interface (eth0 and eth1) and another
subnet for VIPA.
This is a good configuration -- maximum opportunity for redundant
pathways. Depending on the routing setup outside Linux though, you
could have eth0 and eth1 in the same subnet without loss of function
(you would need to make sure that the neighbouring router was setup
redundantly using VRRP, HSRP or equivalent and your OSAs were attached
to different switches).
Why z/OS TCPIP can work with VIPA using two real interface on the same
subnet???
I don't know of any reason why Quagga couldn't do this. See above -- as
long as you ensured that with your interfaces in the same subnet you had
no single-point-of-failure, you'd be fine. Was there a specific problem
you faced that forced you to configure your Linux OSAs in separate subnets?
What I hope you're *not* suggesting is that the z/OS VIPA is in the same
subnet as the interfaces. While this will work, you are not really
providing an opportunity for OSPF to provide you with redundant pathways
to your VIPA [1].
What about Linux on Intel?
Quagga works the same no matter what platform it is built on (subject to
the capabilities of the network hardware of course).
Cheerio,
Vic Cross
[1] Digression: That configuration would possibly work better without
any OSPF at all, by just let the neighbouring router ARP to find your
VIPA. I have not tested or even set up such a configuration, and I
believe it is still IBM's recommendation that VIPAs be in a separate
subnet advertised using RIP or OSPF...
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
----------------------------------------------------------------------
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
