Comments on: Understanding Redistribution (Part I) http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/ Helping you become a Cisco Certified Internetwork Expert Tue, 07 Feb 2012 02:21:38 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: varun http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-165502 varun Sun, 28 Nov 2010 19:46:24 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-165502 Thanks a lot for the information. Cheers varun Thanks a lot for the information.

Cheers
varun

]]> By: KishSquared http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-126984 KishSquared Sun, 08 Aug 2010 12:38:03 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-126984 Josh, Petr is actually correct. The reason the RIP AD must be greater than other externals is because the border router in question is as much a part of the other routing protocol as RIP. For example, R4 above is an OSPF router as much as a RIP router. You want RIP's external routes to be higher than OSPF's external routes. However, RIP doesn't allow this configuration since it lacks the notion of external routes. So the trick to doing this is to make a blanket statement ("all RIP routes are external"), and configure this on the border router. Next, make an exception to the rule by manually setting internal RIP routes below external ADs. This effectively creates the concept of external RIP routes, since RIP doesn't take care of that for us. Finally, Petr was flowing with his train of thought regarding the AD vs Metric. He obviously opted to use metric, but the point is that the internal routers need to somehow "know" which routes are external, and you do that by setting a high metric. You could technically perform AD manipulation on them as well, I believe, but there's no reason to go to such effort unless your RIP domain is huge. Josh,

Petr is actually correct. The reason the RIP AD must be greater than other externals is because the border router in question is as much a part of the other routing protocol as RIP. For example, R4 above is an OSPF router as much as a RIP router. You want RIP’s external routes to be higher than OSPF’s external routes.

However, RIP doesn’t allow this configuration since it lacks the notion of external routes. So the trick to doing this is to make a blanket statement (“all RIP routes are external”), and configure this on the border router. Next, make an exception to the rule by manually setting internal RIP routes below external ADs. This effectively creates the concept of external RIP routes, since RIP doesn’t take care of that for us.

Finally, Petr was flowing with his train of thought regarding the AD vs Metric. He obviously opted to use metric, but the point is that the internal routers need to somehow “know” which routes are external, and you do that by setting a high metric. You could technically perform AD manipulation on them as well, I believe, but there’s no reason to go to such effort unless your RIP domain is huge.

]]> By: Josh http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-115932 Josh Thu, 24 Jun 2010 02:48:33 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-115932 I am confused about something for Rule 1. Petr states "First we should ensure that RIP AD is always greater than any other protocols external AD – on border routers, where this is needed. Next we need to configure so that RIPv2 internal routes have AD less that any other protocols external AD." Shouldn't this read "First we should ensure that RIP AD is always LESS THAN any other protocols external AS"? ie RIP AD = 109 and OSPF external AD = 110 on border router??? Also shouldnt next sentance read "Next we need to configure so that RIPv2 internal routes have METRIC less than any other protocols external routes METRIC" ???? I am confused about something for Rule 1.
Petr states

“First we should ensure that RIP AD is always greater than any other protocols external AD – on border routers, where this is needed. Next we need to configure so that RIPv2 internal routes have AD less that any other protocols external AD.”

Shouldn’t this read “First we should ensure that RIP AD is always LESS THAN any other protocols external AS”? ie RIP AD = 109 and OSPF external AD = 110 on border router???

Also shouldnt next sentance read “Next we need to configure so that RIPv2 internal routes have METRIC less than any other protocols external routes METRIC”
????

]]> By: eric http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-108790 eric Thu, 06 May 2010 02:54:45 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-108790 In this scenario, what if BGP was used to redist the routes? I have a scenario of: OSPF Network AS1/ R1 R2 R3 OSPF Network AS2 I want to run: BGP between R1 and R2 (AS1) OSPF between R2 and R3 (AS2) Can I just redist R1 OSPF into BGP to have the routes on R2, then on R2 redist BGP into OSPF to have the routes on R3? In turn, I was thinking to redist R2 OSPF ingo BGP to get routes to R1, but I think I will hit the loop case. Will BGP prevent the loop?? Or do I still need to do Rule 1 and Rule 2? Thanks!!! In this scenario, what if BGP was used to redist the routes?
I have a scenario of:

OSPF Network AS1/ R1 R2 R3 OSPF Network AS2

I want to run:
BGP between R1 and R2 (AS1)
OSPF between R2 and R3 (AS2)

Can I just redist R1 OSPF into BGP to have the routes on R2, then on R2 redist BGP into OSPF to have the routes on R3? In turn, I was thinking to redist R2 OSPF ingo BGP to get routes to R1, but I think
I will hit the loop case. Will BGP prevent the loop?? Or do I still need to do Rule 1 and Rule 2?

Thanks!!!

]]> By: swallow09 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-23914 swallow09 Tue, 17 Feb 2009 05:52:41 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-23914 !!Thank!! !!Thank!!

]]> By: franc_maurer http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-17997 franc_maurer Tue, 06 Jan 2009 09:29:06 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-17997 I think the Rule 2, though good for loop avoidance, breaks redundancy requirement in most cases (if there is any). In such a simple topology, everything will be fine (redundancy will be maintaned) but consider the situation if R1 is connected with R2 and R3 with point2point links (instead of the shared medium) and R3 has a RIP neighbor directly connected, lets say R10(that router is running only RIP). In such case we should redistribute on R3 all the RIP prefixes learned from R10 and than redistribute them back to RIP from OSPF on R2 (which is in contrast with Rule 2). If we do not do this in if the link R1-R3 fails R1 will have no knowledge about R10 prefixes. You can take even the comparison to OSPF areas further: this is exactly what happens in OSPF if the Area0 is partitioned (because the Area 0 prefixes are not redistributed back to Area 0 if the Are0 is partitioned the reachability is broken) What do you guys think about it? I think the Rule 2, though good for loop avoidance, breaks redundancy requirement in most cases (if there is any). In such a simple topology, everything will be fine (redundancy will be maintaned) but consider the situation if R1 is connected with R2 and R3 with point2point links (instead of the shared medium) and R3 has a RIP neighbor directly connected, lets say R10(that router is running only RIP). In such case we should redistribute on R3 all the RIP prefixes learned from R10 and than redistribute them back to RIP from OSPF on R2 (which is in contrast with Rule 2). If we do not do this in if the link R1-R3 fails R1 will have no knowledge about R10 prefixes.

You can take even the comparison to OSPF areas further: this is exactly what happens in OSPF if the Area0 is partitioned (because the Area 0 prefixes are not redistributed back to Area 0 if the Are0 is partitioned the reachability is broken)

What do you guys think about it?

]]> By: Weiborao http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-14756 Weiborao Wed, 03 Dec 2008 01:38:23 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-14756 Thank Petr Lapukhov, for giving us quite a good article. I have a question about the detail of route redistribution. Is there any Inter-process communication between the two routing processes? For example, a router is running both RIPv2 and OSPF,and redistributing RIPv2 to OSPF.There must be a IPC between RIPv2 and OSPF,right? And from the article "Redistributing Routing Protocols" on CCO. I learned that "The rules for redistribution on a Cisco router dictate that the redistributed route be present in the routing table". So, although the RIPv2 gets updates ervery 30 seconds, the RIP routing table doesn't change, and the redistribution will not be affected.Is that right? Can someone give a deep explanation on route redistributing? Thank Petr Lapukhov, for giving us quite a good article.
I have a question about the detail of route redistribution.
Is there any Inter-process communication between the two routing processes? For example, a router is running both RIPv2 and OSPF,and redistributing RIPv2 to OSPF.There must be a IPC between RIPv2 and OSPF,right?
And from the article “Redistributing Routing Protocols” on CCO. I learned that
“The rules for redistribution on a Cisco router dictate that the redistributed route be present in the routing table”. So, although the RIPv2 gets updates ervery 30 seconds, the RIP routing table doesn’t change, and the redistribution will not be affected.Is that right?
Can someone give a deep explanation on route redistributing?

]]> By: Janardhana Achladi http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-9636 Janardhana Achladi Sun, 12 Oct 2008 02:42:08 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-9636 I got this now. I believe petr intent to explain OSPF and EIGRP ability to distinguish between external and internal routes, where as RIP doesn't. I got this now. I believe petr intent to explain OSPF and EIGRP ability to distinguish between external and internal routes, where as RIP doesn’t.

]]> By: Janardhana Achladi http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-9635 Janardhana Achladi Sun, 12 Oct 2008 02:19:02 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-9635 "For now, we should note that two protocols - OSPF and EIGRP - offer capability to assign different administrative distance values to internal and external prefixes, thanks to their property to distinguish internal and external routes." I did not understand how OSPF will have different AD for external prefixes? Can some one explain? “For now, we should note that two protocols – OSPF and EIGRP – offer capability to assign different administrative distance values to internal and external prefixes, thanks to their property to distinguish internal and external routes.”

I did not understand how OSPF will have different AD for external prefixes? Can some one explain?

]]> By: IE Blog: understanding redistribution « Just another CCIE http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-2748 IE Blog: understanding redistribution « Just another CCIE Sun, 18 May 2008 21:13:48 +0000 http://blog.ine.com/2008/02/09/understanding-redistribution-part-i/#comment-2748 [...] Part I Part II Part III [...] [...] Part I Part II Part III [...]

]]>