ospf loop avoidance

[amms68958]

Hi,

Just wanted to know what is the loop avoidance mechanism used in ospf ..If u point to any url ,it will be of great help..

Thanks in advance..

[amms68958]

Hi all,

Thanks for ur responses....

As maria asked,this scenario is an imaginary one...I used this just to understand whether ,routing loops can occur in that case.(But now I know that my design lacked some basic fundamentals..)

Coming to the design part,if u have any documents which suggests the best practices in ospf design,it will be of great help.

Thanks...

[Kevin Dorrell]

There is a document that everyone quotes, which is the OSPF Design Guide. Here is a link:

http://www.cisco.com/en/US/tech/tk365/technologies_white_paper09186a0080094e9e.shtml

Kevin Dorrell

Luxembourg

[marikakis]

I can recommend the "OSPF Network Design Solutions" book from cisco press.

I consider it the best if someone is really interested in what to avoid in a real network.

Generally, in networking it is good to keep your design simple,

because not only is it easier to implement, but also to troubleshoot.

If something seems very complicated, it is probably not good.

Patches in the design will prove problematic in the long run.

Many stuff and features are better left to the lab and to certification exams ;-)

Of course simplicity should come along with scalability.

M.

[marikakis]

There is one thing I would like to add about partitioned non-backbone areas.

According to a book by Mr. Moy himself ( OSPF RFC author ),

two physically separate regions with the same area Id

will form two separate areas.

The behavior with summaries does exactly that.

It treats the other partition as if it was another non-backbone area.

If the separate regions become physically interconnected,

they will then be able to form a single non-backbone area.

I would like to thank everybody here:

The person who opened this discussion.

Harold for turning the discussion into a useful OSPF lesson.

Kevin for giving me the chance to clarify an answer of mine,

and for his eagerness in learning.

Don't ever think you are "snooping the conversation" ;-)

Best Regards to all,

Maria

p.s. The Moy's book I mentioned is a C++ implementation of the protocol.