Sometimes locally generated traffic doesn't properly hit outbound
ACLs/policies.  Did you verify that you saw matches (counters increasing) on
the EIGRP traffic class?

Also, how were you matching the traffic, match prot eigrp, or with an ACL.
If using an ACL, make sure that you are matching both the destination of
either 224.0.0.10 or the neighbor's address.

The CCIE lab is full of situations where you can be asked to do a normal
thing, but then told to not do it a certain way.

On a side note, the "ip bandwidth-percent eigrp" is a very interesting
command, because it is a percentage command that will allow you to specify a
number greater than 100, which could be used if the bandwidth on the
interface was set to a lower value than what the circuit actually was.

Just curious, is there a reason why you chose policing over shaping?

Marvin Greenlee, CCIE #12237 (R&S, SP, Sec)
Senior Technical Instructor - IPexpert, Inc.
Telephone: +1.810.326.1444
Fax: +1.810.454.0130
Mailto: [EMAIL PROTECTED]

Progress or excuses, which one are you making?



-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Tony Hidalgo
Sent: Thursday, July 10, 2008 4:46 PM
To: [email protected]
Subject: [OSL | CCIE_RS] EIGRP

Hello people from the list.

On the "Focus Labs", Section 9 (EIGRP), question 9.18 it is requested to set
the EIGRP bandwidth of a FR link to 37.5%. This WITHOUT using an interface
based command (that would be the ip bandwidth eigrp AS# %).

The PG gives a funky solution of actually changing the BW of the interface
itself. I frankly disagree with that answer although it may accomplish the
goal from some perspective.

The solution that I thought of was MQC. I created an ACL to match eigrp
traffic. Then a policy map to "police cir 579000" (579K). This because the
BW of the interface is 1544Ks (default) and this represents the 37.5% of the
total BW of the FR interface. Then, I applied the policy map OUTBOUND on the
interface in question.

Since I am not breaking any rules or requirements, does this look like a
valid solution??

THX



      

Reply via email to