10 - Traffic Engineering Flashcards Preview

CS 6250 Test 3 > 10 - Traffic Engineering > Flashcards

Flashcards in 10 - Traffic Engineering Deck (20)
Loading flashcards...

What is Traffic Engineering?

Traffic engineering is a way to describe how companies deal with a large amount of data flowing through their networks.

It is also the process of reconfiguring the network in response to changing traffic loads to achieve some operational goal.


What are some of the reasons a network operator may want to use traffic engineering?

- To Maintain Peering Ratios

- To Relieve Congestion

- To Balance Load Across Links


What are some examples of IP self-management?

- TCP Congestion Control

- Routing


What is the problem with IP self-management?

The network may not run as efficiently as possible?


What is the key question that traffic engineering tries to address?

How should routing adapt to traffic?


In terms of adapting routing, what does traffic engineering attempt to do?

- avoid congested links

- satisfy application requirements


How might an operator route traffic off of a congested link?

By increasing the link weight.


What are some of the ways that operators set link weights?

- Inversely proportional to capacity

- Proportional to propagation delay

- Network-wide optimization


What are the 3 steps of traffic engineering?

1. Measure - figure out current traffic loads

2. Model - of how configuration affects the paths in the network

3. Control - reconfigure to control how traffic flows


What is the input to the problem of solving Intradomain Traffic Engineering Optimization?

1. Graph G (Routers, Links) and Capacity of Links

2. Traffic Matrix Mij - the traffic load from outer i to router j


What is the output of the problem of solving Intradomain Traffic Engineering Optimization?

A set of link weights where Wl is the weight on any unidirectional link l.


What are some practical solutions to the fact that no algorithm exists to solve the Objective Function of minimizing traffic on all links?

We have to resort to searching through a large combination of link weight settings.

While this is sub-optimal, the graphs are often small enough that this approach is reasonably effective.


In light of the un-solvable Objective Function, what are some other operational realities that have to be considered when implementing a solution?

- The solution should minimize the number of changes to the network. Often just changing one or two link weights is enough to achieve a good traffic load balance solution.

- Solution must be resistant to failure

- Should be robust to measurement noise


What is the difference between Intra vs Interdomain Routing?

Intradomain: Within a domain (ISP, campus, or datacenter)

Interdomain: Between domains


How is Interdomain Traffic Engineering implemented?

Through reconfiguration of the BGP (Border Gateway Protocol).


What are some goals of Interdomain Traffic Engineering?

- Alleviating congestion on edge links

- Using new/upgraded edge links

- Changing end-to-end paths


What are the goals of effective Interdomain Traffic Engineering?

1. Predictability

2. Limit influence of neighboring domains

3. Reduce overhead of routing changes


What is one way to achieve predictability in Interdomain Traffic Engineering?

Avoid making route changes that are globally visible?


What is one way of using Interdomain Traffic Engineering to limit the influence of neighboring domains?

- Treat paths that have *almost* the same AS path length as a common group to achieve additional flexibility.

- Enforce a constraint that neighbors advertise consistent BGP route advertisements over multiple pairing links.


How can Interdomain Traffic Engineering be used to reduce the overhead of routing changes?

Group related prefixes. Rather than exploring all combinations of prefixes to move a particular volume of traffic, we can identify routing choices that group routes that have the same AS paths.