I.1) presentation of the document
- Part of my teenager life passed-up through the 1st Internet revolution (in 1997, with AltaVista, Yahoo, etc.), and as far as I am concerned, the Internet has changed our way of life and impacts everything everyday. Even my country which is Djibouti (“horn of the East-Africa”), Internet was already provided since 1996, and began to invest in submarine cable infrastructure for planning to position itself later as a regional Internet hub.
BGP mechanisms allow us to understand deeply how the Internet works at a technical view.
- As a result, my first concern about writing this document is to share my knowledge of BGP that I have, and received from others BGP operators and experts.The document does not have the pretension to propose the best and accurate text, but only makes part so far, of my work, and invites other people to share their view and their understanding.
- This document is destined to readers who have a reasonable background in corporate and Internet networking, but are not experts in BGP or backbone Internet routing architectures.
- The target audience for this text concerns mainly networks administrators and engineers who wish to get a familiarization with BGP, in an applicable way, on how to process IP transit connectivity. For instance, administrators who have passed their CCNP and CCIE recently, will find this document interesting to a practical point of view. Still, certifications are not an obligation; it is indicated for reference only.
- Administrators or experts, who manage considerable networks size every day for large ISPs, can use also this document for their trainees, lol.
How to use this document;
- The aim is to give a demonstration on how to use BGP and understand it in a world of interconnected ISPs.
The contents describe practices and techniques about an ISP customer who:
- buys IP transit connectivity from different ISPs.
- needs to implement a load-balancing process.
- inserts redundancy and multihoming options for the Internet provision in case of link failure.
- You will not find all the knowledge about BGP, only the essentials on what is necessary to the purposes described above.
- The tutorial and situations have been simplified as much as possible; information is accurate and precise, has been tested in a real ISP, and also through a simulator as much as I could.
- The document displays many uses of Cisco command and outputs; this is because my work, and my technical choices make me practice mainly with Cisco’s equipment only. Even the design of the diagrams, feels like a Cisco project implementation, this is due to my formation. Nevertheless, there is not any advertising goal about Cisco’s technologies, only standards protocols are described and discussed.
- The chapters have been organized in a way that you can jump from one to another one without any restrictions too. Some situations described, are also deliberately complicated, this in order to highlight a result or emphasis a notion. Some readers can tell why complicating things in such a way, this is only for an informative goal. Of course, it is possible to optimize the configuration.
- The paper is composed of four chapters:
- The first chapter introduces the design of the network and its topology for the study case.
You will find diagrams on the BGP and Static design; the traffic engineering targeted configuration, and a listing of IP-addresses and interfaces used.
- The second chapter (chapter A) highlights various global notions on BGP, which are a necessity to understand. This is in order to be able to design and configure the network described in this study case, which aims to deliver IP transit services to customer, and uses multihoming connectivity to several ISPs.
The general points of BGP in this chapter are related to a way of thinking within an ISP environment. People can overview it, but it is recommended to read it accurately if they do not practice BGP administration every day.
- The 3rd chapter (chapter B) targets the design directly and its configuration. Ebgp and Ibgp configurations are described and deeply explained. Peering agreements and multihoming are enlightened in a practical way.
- The last chapter (chapter C) validates the configuration of ChapterB, and more generally the study case. Scenarios of links failures are presented, in order to check up if the multihoming configuration agrees the design and the traffic engineering objectives. Moreover, verifications, monitoring and debug commands are also stressed, in order to validate the setup.
- To conclude, with all the applicable and practical information and contents in this document, readers can easily manage, improve or figure a Tier3 or Tier4 ISP backbone Internet network.
- All best practices have not been included, only the one which focus on the routing and verification procedures. For common best practices, other experts have already done an excellent work. Network operators group like MENOG, NANOG, AFNOG, etc. give also much useful information on these topics.
- At the end, I would like to thanks:
- the Internet organisms and its founders, and everyone promoting the Internet structure to remain the same: free at 100%.
- all network engineers who share on the Internet their knowledge about BGP and others networks studies.
- all my friends, colleagues, work relations who encourage me to write this tutorial and review it.
- English is not my native language, so you will excuse me for the expression.
I also invite readers to criticize my work technically (in other words, me and whatever) and visit the following websites and resources that you can find at the end of the document.