It's not a dumb question, as far as I can tell, but it's awfully hard to
parse due to your stream of consciousness style and lack of paragarpahs.
White space is a good thing. :-)

There's nothing wrong with IP secondary addresses as far as I know. They can
come in handy. And yes you can have multiple ones. At some point you should
design your network more hierarchically and get rid of them perhaps, but
that's for another thread.

The one caveat is that packets in and out the same interface, for the few
cases that you mentioned that they do communicate (through this interface?),
you need to enable fast switching. It's not on by default. And.... hmmmm...
I can't remember the command. Maybe somebody else will remember it.

The rumor that secondary addresses are bad has to do with IPX. When
supporting more than one encap for IPX you can do it with secondary or
subinterfaces. Cisco said they were going to discontinue support for
secondary. I don't think they ever did. And they caused lots of confusion
because people assumed they were talking about IP too, which they weren't.

In the IP world, secondary addresses can be very useful, especially for what
you're describing where you have two sets of addresses on one LAN, and maybe
more than two in the future.

Subinterfaces are for VLANs as mentioned in other messages (and IPX multiple
encaps.)

Priscilla


Mark Smith wrote:
> 
> I have a router (actually a pair of them in HSRP but that's
> irrelavent) that
> connects two networks in non-contiguous IP address ranges
> through a 100MB
> F/E TX port on the inside to an ISP network on the outside,
> also via a 100MB
> F/E TX port, at a colo facility. I have a half of a hundred MB
> pipe to the
> outside world. The two networks behind the router are
> independent of each
> other, seperated by PIX's behind the routers but, on occasion,
> they do
> communicate with each other. I currently have a primary and
> secondary IP
> address set on the inside F/E interface, one for each network. 
> I've never
> seen any mention if sub-interfaces being used in Ethernet or
> Fast Ethernet
> interfaces in any Cisco literature. They primarily seem to be
> mentioned in
> regards to serial interfaces. Is there an advantage to using
> sub-if's here
> over a primary and secondary IP address? Any packet filtering
> is handled by
> the PIX's so I don't ever foresee the use of access-lists on
> the router.
> This router simply routes packets. I don't foresee the use of
> more than two
> networks inside but I suppose that's a slight possibility down
> the road if I
> would need more IP addresses and couldn't get contiguous
> addresses.  I'm not
> sure if you can use more than a single secondary address on an
> interface or
> if you can pnly use a single one. I guess I'm not sure if
> "recommended
> practice" would be to always use sub-if's when connecting more
> than one
> network to any interface, use sub's only with serial i/f's and
> use
> primary/secondary addresses with F/E interfaces or if it's time
> to consider
> adding more F/E modules with 2 or more networks. I've used this
> primary/secondary config for a couple of years and it's worked
> fine but, as
> my colo facility is Sprint and they've decided to get out of
> the hosting
> biz, it would be a good time for me to reconfigure things
> during the move if
> there is an advantage in doing so.
> Thanks.
> 
> 
> 
> 




Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=71461&t=71447
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]

Reply via email to