Static routes vs. routing protocol

Static routes vs. routing protocol

Post by Bill Kie » Tue, 04 Mar 1997 04:00:00



Can anyone explain to me exactly now static routes can save on
overhead of the router.  I know if you using a routing protocol that
it keeps a routing table and updates are sent to the other routers,
but when you use static routes I assume that it acts like a switch and
just send the packets to where they need to go based on the static
route....and it doesn't broadcast anything to the other routers.  This
is going to be used over a frame-relay network using subinterfacing
(frame-relay interface-dlci xxx).  My thought is that I can save a lot
of overhead because the remote sites are just a small ethernet segment
and none of the remote sites need to talk to each other.  

 
 
 

Static routes vs. routing protocol

Post by Jeff Doyl » Wed, 05 Mar 1997 04:00:00



>Can anyone explain to me exactly now static routes can save on
>overhead of the router.  I know if you using a routing protocol that
>it keeps a routing table and updates are sent to the other routers,
>but when you use static routes I assume that it acts like a switch and
>just send the packets to where they need to go based on the static
>route....and it doesn't broadcast anything to the other routers.  This
>is going to be used over a frame-relay network using subinterfacing
>(frame-relay interface-dlci xxx).  My thought is that I can save a lot
>of overhead because the remote sites are just a small ethernet segment
>and none of the remote sites need to talk to each other.  

Bill:

Static routes can save both on router processing and on bandwidth.  

When a router receives updates, it has to run an algorithm to determine
the best path and then place that route in the routing table (this
is an extremely simplistic description of what actually happens).  The
algorithm may be a very simple one, such as what RIP uses, or very
complex, as in OSPF.  The algorithm obviously translates into CPU
cycles, with more complex algorithms calling for more CPU.  Static
routes do not require these algorithms-- *you* are the algorithm, telling
the router where to send packets.

Whether you use dynamic or static routes has no impact on performance
during actual routing-- the routing table lookup just looks for whatever
routes are in the table, regardless of how the route entries got there.

Where static routes *really* save, IMHO, is in bandwidth.  RIP broadcasts
the routing table every 30 seconds, OSPF sends LSAs, etc.  If there are
no decisions for a routing protocol to make, why use up some of your
bandwidth with what is essentially useless information?

For instance:  Small frame relay networks are usually set up in a
hub-and-spoke topology.  If one of those spokes dies, can the
router at the end of it find an alternate path back to the hub?  If
not, what use is a routing protocol, other than to use up some percentage
of your expensive WAN bandwidth.  

In such a topology, configure static routes in your hub router pointing
to each LAN on the spoke routers, via the frame relay interface addresses
of the spoke routers.  On the spokes, add a single static default route:

IP ROUTE 0.0.0.0 0.0.0.0 A.B.C.D

where A.B.C.D is the frame relay interface of the hub router.  Add
IP CLASSLESS to all routers, and you're in business.  No muss, no fuss.

Jeff
--------------------------------------------
Jeff Doyle
Customer Service Specialist
CCIE # 1919
Certified Cisco Systems Instructor

!nterprise Networking Services
US West Communications, Inc.
201 Third St. NW, Rm. 720              
Albuquerque, NM  87102

          Office:  (505) 245-5857
             Fax:  245-6189


      Compuserve:  74051,656
---------------------------------------------

 
 
 

Static routes vs. routing protocol

Post by Lance Ecklesdafe » Fri, 07 Mar 1997 04:00:00


Bill,

We are dealing with the exact same question.

I would love to hear about any responses you get to your question.  We are
suing EIGRP currently. The sites are not talking to each other and there
are no other routers present at any of these remote sites. My thought is
that we could even use IP UNNUMBERED interfaces for the serial llinks over
frame-relay.

I have had this conversation with others and they seemed to think that it
would work fine.

Lance Ecklesdafer
Perot Systems Corporation



Quote:> Can anyone explain to me exactly now static routes can save on
> overhead of the router.  I know if you using a routing protocol that
> it keeps a routing table and updates are sent to the other routers,
> but when you use static routes I assume that it acts like a switch and
> just send the packets to where they need to go based on the static
> route....and it doesn't broadcast anything to the other routers.  This
> is going to be used over a frame-relay network using subinterfacing
> (frame-relay interface-dlci xxx).  My thought is that I can save a lot
> of overhead because the remote sites are just a small ethernet segment
> and none of the remote sites need to talk to each other.  

 
 
 

Static routes vs. routing protocol

Post by Barton F. Bru » Sat, 08 Mar 1997 04:00:00



> Bill,

> We are dealing with the exact same question.

> I would love to hear about any responses you get to your question.  We are
> suing EIGRP currently. The sites are not talking to each other and there

  ^^^^^

Of course you mean "sueing" :-)

I too have considered that after many EIGRP + frame relay 'experiences'.

 
 
 

Static routes vs. routing protocol

Post by Lance Ecklesdafe » Sat, 08 Mar 1997 04:00:00


Yes I mean sueing ... of course who do you sue for this protocol? -LOL
Of course there is the question of why I do not spell check posts to usenet
:)

Lance





> > Bill,

> > We are dealing with the exact same question.

> > I would love to hear about any responses you get to your question.  We
are
> > suing EIGRP currently. The sites are not talking to each other and
there
>   ^^^^^

> Of course you mean "sueing" :-)

> I too have considered that after many EIGRP + frame relay 'experiences'.

 
 
 

1. static routes vs a routing protocol

While this is not a cisco specific issue, I hoping that someone on this
list can help me out.  I'm looking for any reference material that covers
the pro's and cons of employing a routing protocol vs. static routes
especially on stub networks.

I recently accepted a position as part of a statewide project to bring low
cost internet access to k-12 education, libraries, mueseums, etc.  And
while there are currently in excess of 1000 sites with connections into the
network, the latitude given to individuals responsible for the different
regions of the state has led to an adhoc approach to routing on this
network.  Currently, a seemingly random mix of eigrp groups and static
routes are employed between the distibution and access routers.  

I would like to see us take a uniform approach on how routing is handled
between the these routers but I need to be able to justify my
recommendation.  

thanx,
-john

2. FTP problems

3. Static route vrs. Routing protocols.

4. Windows 2000 Technical Question

5. Routing Protocol Support vs. Routed Protocls

6. Update Sucess

7. Router on x25 static route Network does not know second static rout is unavalible

8. SPARCstation 20 135mhz f.s.

9. Can netwrok run static route and dynamic route the same time?

10. Static routes AND policy routing?

11. static routing and route-map

12. How static routes come in & out of the routing table

13. Floating Static Routes and unnumbered routing