|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
From: Ash he (ash_he
YAHOO.COM)Date: Thu Aug 23 2001 - 15:54:38 CDT
Hi Xie,
Will you not need to store multiple paths to
implement equal-cost routing or traffic-load
distribution?
-Ash
--- "Xie, Feng" <Feng.Xie
MARCONI.COM> wrote:
> Hi, Nikhil:
>
> It is still not very clear to me why we need
> to store multiple paths
> to an ASBR. Please correct me if I am wrong. My
> understanding is that an
> ospf router only needs to keep the best route to an
> certain ASBR according
> to the rules specified in RFC2328 section 16. In
> implementation, you can
> choose to keep all routes to an ASBR and then choose
> the best one every time
> an external prefix is processed in SPF calculation.
> Or you can only keep the
> best route to an ASBR. Every time you receive a
> different route to the same
> ASBR, you compare it with the one you keep. If
> better, replace the old one.
> For me, the second one seems better.
>
> It is possible that multiple ASBRs advertise
> the same external prefix.
> Since there is a best route to each ASBR, you do
> need to use the spec in
> Section 16 to choose the best among them.
>
> Feng Xie
>
> -----Original Message-----
> From: Nikhil Sama [mailto:Nikhil.Sama
NEXSI.COM]
> Sent: Thursday, August 23, 2001 1:58 PM
> To: OSPF
DISCUSS.MICROSOFT.COM
> Subject: Re: About ASExternal calculation
>
>
> Alex,
> Actually you do keep multiple routes to a
> router destination.
> For instance, if an ASBR is reachable in
> multiple area's via
> different paths, you must store a routing table
> entry for each path
> regardless of cost.
> It is only in the case of IP routes that you
> store only the "best
> route/routes" to a destination.
> Nikhil
>
> "Xie, Feng" wrote:
>
>
> Hi, Alex:
>
> Usually, we only keep the best route to a
> destination. As you said, at
> any time, there may be only one entry for an ASBR
> with multiple next hops
> due to equal-cost multipath. My understanding is
> that what described in the
> RFC is a process to build the best route to an ASBR
> when you have choice
> among multiple routes. It does not mean that an
> implementation has to keep
> all possible routes to an ASBR.
>
>
> Feng Xie
>
>
> -----Original Message-----
> From: alex lerin [ mailto:alexlerin
YAHOO.COM
> <mailto:alexlerin
YAHOO.COM> ]
>
> Sent: Thursday, August 23, 2001 5:43 AM
> To: OSPF
DISCUSS.MICROSOFT.COM
> Subject: About ASExternal calculation
>
>
> Hi..
> I have a doubt in ASExternal Calculation...
>
>
> In ASExternal calculations of RFC2328, Section
> 16.4,3rd point 3rd paragraph,it says that if the
> forwarding address is zero,th packet should be sent
> to
> the ASBR itself.Also it says that we have to prune
> for
> multiple entries.
> My doubt is that how can we prune when there is
> always only one entry for the ASBR in the routing
> table.Because whenever we get a destination ,we
> check
> for its entry in the routing table and compare the
> path type etc.,and if thay are equal cost we only
> add
> the next hop entry to the route and not a different
> entry.So there is no possibility for multiple
> entries
> for the same ASBR in the routing table.
> If I am wrong tell me when can there be more than
> one entries for the same ASBR in the routing table?
>
>
> Thanks in advance
> Alex
>
>
> __________________________________________________
> Do You Yahoo!?
> Make international calls for as low as $.04/minute
> with Yahoo! Messenger
> http://phonecard.yahoo.com/
> <http://phonecard.yahoo.com/>
>
> --
>
> Nikhil Sama E-Mail:
> Nikhil.Sama
nexsi.com
>
> Routing Protocols Group Ph: 408 579 5880
>
> Nexsi Corp. Fax: 408 579 5701
>
> 1959 Concourse Drive Ph: (M) 310 428 9191
>
> San Jose, CA 95131. www.nexsi.com
>
>
>
> Home:
>
> 1240 Dale Av., Apt 16 Ph: 650 237 9055
>
> Mountain View, CA 94040
>
>
>
__________________________________________________
Do You Yahoo!?
Make international calls for as low as $.04/minute with Yahoo! Messenger
http://phonecard.yahoo.com/
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]