From: Howard C. Berkowitz (hcb@gettcomm.com)
Date: Thu Jul 15 2004 - 15:37:41 GMT-3
At 9:58 AM -0700 7/15/04, Joseph D. Phillips wrote:
>There are many political or economic reasons why optimal routing isn't
>always possible or desired.
>
>The job of a network engineer, as I see it, is to ensure data flows
>according to the requirements and directives of the organization. You
>may have great ideas about how to improve the flow of traffic, but you
>can slam into bureaucratic walls trying to implement your designs.
>
>Can I get a witness? Scott? Brians? Howard?
Thanks -- I've been meaning to respond.
Optimal routing has no single meaning. For example, there is the
concept variously called hot potato or closest exit routing. Examples
here would be an ISP handing off the packet to the first exit it can
find, so the workload imposed on the ISP backbone is minimized.
In like manner, in an OSPF totally stubby area with multiple ABRs,
you only have information to get the packet to the closest ABR -- you
are only locally optimizing within the area. Broadening the scope of
OSPF to ISPs, you can use Type 1 externals to load share within your
domain, but, without traffic engineering, you have no control of
optimization once the packet leaves your domain.
It's been my experience that people tend to overemphasize routing by
metric, etc., in that there usually aren't so many different possible
routes that the best metric really makes much different. Topology,
particularly hierarchical topology, tends to be far more important.
I'm working on some practice labs in which I consciously have a
section called "the right way, the wrong way, and the Cisco way." In
the one I'm testing at the moment, the basic solution will tend to
put traffic over a 64 Kbps link known to OSPF, rather than a 10 Mbps
link known to Ethernet. I have a specific last step that plays games
with redistribution metrics and distances to force the traffic onto
the "best" path. It is my impression that Cisco would not ask for
such in the CCIE lab, and I'm trying to emphasize the difference --
trying to tell people when something is good enough for the lab
requirmeents.
>
>
>
>
>
>----- Original message -----
>From: "Kenneth Wygand" <KWygand@customonline.com>
>To: "Kenneth Wygand" <KWygand@customonline.com>, ccielab@groupstudy.com
>Date: Thu, 15 Jul 2004 12:40:30 -0400
>Subject: Re: Suboptimal Routing?
>
>Surely someone has an opinion on this or can explain why this is (or
>should be) done.
>
>Thanks in advance,
>Ken
>--------------------------
>Sent from my BlackBerry Wireless Handheld
>
>
>-----Original Message-----
>From: nobody@groupstudy.com <nobody@groupstudy.com>
>To: ccielab@groupstudy.com <ccielab@groupstudy.com>
>Sent: Wed Jul 14 20:57:34 2004
>Subject: Suboptimal Routing?
>
>I have a generic question regarding suboptimal routing on the CCIE lab.
>
>Suppose I have both RIP and OSPF running in my network. I have two
>routers that are running both RIP and OSPF (on the RIP/OSPF border).
>One of these routers is mutually redistributing RIP and OSPF (let's call
>this "Router 1"). Without changing distances or filtering, routes that
>originate in RIP and are redistributed by "Router 1" into OSPF will be
>received by "Router 2" in OSPF. So now Router 2 will know of routes
>from the RIP domain through both it's own RIP routes and also through
>the OSPF routes. Of course, the distance of OSPF is less than RIP, so
>it will default to choose the routes that "go all the way around the
>world" to get back into the RIP domain.
>
>This is a clear case of "suboptimal routing" but still provides full IP
>connectivity. Every single practice lab I've ever come across has
>always "treated" this situation by using some mechanism (filtering,
>distance, etc) to cause the native RIP routes to be preferred through
>RIP as opposed to the "all the way around the world" OSPF routes.
>
>However, none of these labs ever say "ensure you avoid suboptimal
>routing". I've also heard many times that the lab is "not testing
>real-world best practices, only that you meet the requirements of the
>question". Well, if the question states I must have IP reachability,
>then why should I bother fixing suboptimal routing. Sure, it will
>display to the proctor that I understand it and I know what I'm doing,
>but why -specifically- do I have to do it to get points on my exam? If
>it doesn't say to avoid this, why should I waste my time?
>
>Thanks in advance, as I really look forward to your responses.
>Ken
This archive was generated by hypermail 2.1.4 : Sun Aug 01 2004 - 10:11:56 GMT-3