RFC 3345 Persistent Route Oscillation Condition - Lab results and concerns

Unanswered Question
Aug 2nd, 2010

Dear friends,

I've been running some labs, whose topology is the despicted on the Type I Churn scenario, and I got the following results:

- IOS version earlier 2002 (year of RFC 3345 publising) are subjected to that issue (tested on dynamips with 11.2)

- IOS version released after 2002 are not affected to that issue (tested on dynamips with 113-11b.T5)

By searching RFC Editor (www.rfc-editor.org) I found out at the time RFC 3345 publishing RFC 1771 was the most updated one regarding BGP. In 2006 the RFC 4271 was released obsuleting 1771. I was able to realize that section Breaking Ties (9.1.2.1 at older one and 9.1.2.2 at the newer one) quitly changed. I think the persistent route oscilation stated on RFC 3345 was broken due to the deployment of the algorithm proposed at c), 9.1.2.2 of RFC 4271. From my point of view that algorithm looks like deterministic MED. Am I correct? If yes, why has IOS mantained the command bgp deterministic-med yet?

I'm very pleased if anyone can drop some comments.

Best regards,

Alberto

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.

Actions

This Discussion