All,
I was looking into a test scenario with a co-worker and we stumbled
upon
something very interested as it applies to the routing theory and
redistribution
of IGRP/EIGRP and RIP. This is the basic setup (hopefully the ASCII art
makes
it..)
172.16.10.32/27
172.16.10.64/27
172.16.10.1/27|--------R1(e0)---------------------------(e0)R2(e1)-----------
----------------(e0)R3----| 172.16.10.96/27
The R1-- R2 link has IGRP(300) and R2--R3 link has EIGRP(300). On R2 I add 2
loopbacks on the following networks 10.0.0.0/8, 170.10.0.0/16, and
192.168.10.0/24. These 3 loopbacks are then added to a RIP process on R2.
The
RIP process is then redistributed to
into the EIGRP process on R2 as follows;
router eigrp 300
redistribute rip
network 172.16.0.0
no eigrp log-neighbor-changes
!
router rip
network 10.0.0.0
network 170.10.0.0
network 192.168.10.0
!
router igrp 300
redistribute rip
network 172.16.0.0
Now notice on redistribution of RIP into the EIGRP I did not use any metrics
nor did I make use of the "default-metric" command. As expected
none of the RIP networks show up on R3. However when I added the
"redistribute rip" to the IGRP process the routes showed up on R1.
So now my question is "How come?". I looked at the default-metric command
and everything says that metrics/or the default-metric
command to inform IGRP/EIGRP on how to handle the routes received from
protocols like RIP, OSPF and ISIS. I've only tested this with
RIP and will conduct further testing to see if the other protocols I
mentioned
behaves the same.
The only thing I could find that could possibly explain what I'm seeing is
referenced by this link... (watch the wrap)
http://www.cisco.com/univercd/cc/td/doc/product/software/ios121/121cgcr/ip_r/
iprprt2/1rdeigrp.htm#xtocid189574
The interesting text is as follows..
Defaults: Only connected routes and interface static routes can be
redistributed without a default metric
Usage Guidelines
A default metric is required to redistribute a protocol into IGRP or Enhanced
IGRP, unless you use the redistribute command.
Automatic metric translations occur between IGRP and Enhanced IGRP. You do
not
need default metrics to redistributed IGRP
or Enhanced IGRP into itself.
The question I have is why does it work for IGRP and not EIGRP? Ok.. here's
a debug on R2..
3d09h: IGRP: sending update to 255.255.255.255 via Ethernet0 (172.16.10.34)
3d09h: subnet 172.16.10.96, metric=1600
3d09h: subnet 172.16.10.64, metric=1100
3d09h: network 170.10.0.0, metric=501
3d09h: network 192.168.10.0, metric=501
3d09h: network 10.0.0.0, metric=501
3d09h: IGRP: Update contains 2 interior, 3 system, and 0 exterior routes.
3d09h: IGRP: Total routes in update: 5
Ok, so as we know RIP and IGRP both use broadcast to update neighboring
routers of route information whereas EIGRP and OSPF use multicast. I can see
that it is possible for IGRP to hear this route information. This still
doesn't answer my first question... How does IGRP know what metrics to apply
to the RIP routes without being told.
Anyone care to shed some light on my otherwise dark mind. Is this expected
behavior?
Nigel.
P.S Debbie you found anything... ?
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=20962&t=20962
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
