A Network Artist : RFC3101

Showing posts with label RFC3101. Show all posts

Sunday, October 19, 2014

OSPF Route Preference: You Sure You Know It ? ... RFC 3101...Part 2 (Final)

It's sunday, got some time today. Let's wrap up this series quickly.

The physical and logical topology is gonna remain same. The only thing I changed in this part is ... IOS Version :)

R3 is Still the Translator:

Let's review the routing table of R2 quickly again and see how things are:

So now the question is why results are different with different IOS ? ( IOS used in part 1 was 12.4 (20) T ).

The previous version was actually following RFC 1587 in which the route preference is same as mentioned in post 1 i.e. E2 is preferred over N2.

But the later IOS followed the RFC3101 implementation. Which states that LSA with P-bit set is preferred over the one which has P bit as Zero. The P bit is set to 1 by ASBR while redistributing the routes into NSSA except the situation where ASBR is also ABR and P bit remains unset (0).

But as an Engineer if you would like you can still go back to older IOS behavior as following:

The reason in Part 1 , the traffic (Data Plane) took the direct path because of same cost to Forwarding Address.

Further Readings:

http://blog.ipspace.net/2008/01/e1-and-e2-routes-in-ospf.html

http://www.cisco.com/c/en/us/td/docs/ios-xml/ios/iproute_ospf/configuration/15-e/iro-15-e-book/iro-ospfv2-nssa-cfg.html#GUID-C335FD95-ED21-4A51-894F-8B63119186EC

http://networklessons.com/ospf/ospf-path-selection-explained/#ixzz3GPsDJ8N8

HTH...
Deepak Arora
Evil CCIE

Saturday, October 18, 2014

OSPF Route Preference: You Sure You Know It ? ... RFC 3101...Part 1

Well I wanted to write about this for quite a while but seem to lost in time. We all know that OSPF is fairly a complex protocol to understand and work on though It makes me wonder why People usually come with OSPF as preferred choice if I ask them to choose IGP for New Enterprise Network Design. Well let's keep that story for my Network Design Series.

Now if you know me, I am one of those people who find reading books cover to cover quite hard. I prefer other methods like Video Trainings, Labing Up the topic I am trying to understand, Cisco Documentation and of course above all the Blog Posts at different forums.

While reading one of those exiting blog posts long back I came across this post from Ethan Banks on an interesting OSPF issue which he discovered during practicing one of Net Master Class CCIE R&S Lab from their workbook.

So I decided to test the scenario by my own thinking I can re-produce the issue and see get it working the way Ethan suggested.

So I created this quick topology (Of course Replaced Cat 1 with R4).

I configured everything like shown in Diagram but to my surprise the Issue didn't occur.

Now the questions was why so ?

I mean from theory standpoint the issue should have occurred. Let's review the topology from OSPF LSA Standpoint and see:

In my topology R3 became the translator.

Now by theory R2 should have seen LSA-5 (Translated by R3 & Forwarded By R1) as well as LSA-7 received directly from R4. In which case R2 should prefer LSA-5 over LSA-7 based on OSPF Route Selection Mechanism:

Regardless of a route’s metric or administrative distance, OSPF will choose routes in the following order:

Intra-Area (O)
Inter-Area (O IA)
External Type 1 (E1)
External Type 2 (E2)
NSSA Type 1 (N1)
NSSA Type 2 (N2)

Now let's review what routing table of R2 tells us first:

Well it seems to be following same logic. Let's take a closer look at Database Table now:

Now control plane seems to be in Sync with what Ethan suggested. But what about Data plane ?. Will R2 send traffic towards R3 to get to 200.200.200.200 network ?