Traffic Engineering

Question 1

What is Traffic Engineering

Answer 1

Process of reconfiguring the network in response to changing traffic loads, to achieve some operational goal

Question 2

What are key application requirements that traffic eng accounts for

Answer 2

Avoid congested links satisfy applicationrequirements

Question 3

Link weight allow?

Answer 3

Operators to control the flow of data trough the network

Question 4

Link weight influences which alogorithm

Answer 4

Shortest path

Question 5

Way link weight are set in practice

Answer 5

Inversely proportional to capacity, Proportional to propagation delay, Network-wide optimization

Question 6

3 steps of traffic engineering

Answer 6

Measure, model, control

Question 7

Input for intradomain TE optimization

Answer 7

Input: Graph G(R, L) R= routers {e: routers}, L = links {e : links}, Ce = capacity of L

Also considered the traffic load matrix Mij

Question 8

What is the output of intradomain optimization

Answer 8

W = {e: link weights}

Question 9

How is utilization defined

Answer 9

UL = amount of traffic on the link, CL = capacity, UL/CL is the amount of traffic divided by capacity

Question 10

Objective of intradomain optimization

Answer 10

Minimize sum of F(UL/CL) I assume they mean find the best parameters to fit an ideal curve (lecture says it is NP-Complete)

Question 11

Since minimizing the sum of utilizations is NP what is done in practice?

Answer 11

• Change a few link weights, Resistant to failure, Robust to measurement of noise, Minimize number of changes to network

Question 12

Intradomain routing

Answer 12

Within a domain (e.g. ISP, Campus, data center)

Question 13

Interdomain routing

Answer 13

Between domains

Question 14

Interdomain Traffic engineering

Answer 14

Alleviate congestion on edge links, Using new/upgraded edge links, Changing end-to-end paths

Question 15

What mechanism is used for interdomain traffic engineering

Answer 15

Reconfigure of BGP

Question 16

Three goals of interdomain TE

Answer 16

Predictability, Limit influence of neighbors, reduce overload of routing changes

Question 17

How is predictability achieved

Answer 17

Avoid globally visible changes that are broadcast outside of an AS

Question 18

How is limiting influence of neighbors achieved

Answer 18

Consider neighbors with a similar path length as a group (I guess we used some minimal distance or distance threshold to determine this?)

Can also enforce consistent advertisements

Question 19

How is reducing overload of routing changes achieved

Answer 19

Group prefixes and move traffic in terms of these prefixes

Question 20

ECMP

Answer 20

Equal cost multipath

Question 21

One way to achieve Traffic spliting

Answer 21

The router could have multiple forwarding entries with different next hops

Question 22

Data center characteristics

Answer 22

1. Multi-tenancy (Pro amortization of cost, Con Security, resource isolation) 2. Elastic resources 3. Flexible service management Workload movement migration

Question 23

What mechanism makes Data center objectives achevable

Answer 23

Virtualization

Question 24

Data center challenges

Answer 24

Traffic load balancing, Support for virtual machine migration, Power saving, Provisioning, Security

Question 25

3 layers of historic data center

Answer 25

Access (servers) are below this, Next above this is Aggregation, and then connected to the internet is the Core

Question 26

Historically data centers used what topology

Answer 26

Layer 3 topology now they use layer 2

Question 27

What are the benefits of Layer 2

Answer 27

Migration, load balacing

Question 28

Draw backs for layer2 topology for data centers

Answer 28

Scale

Question 29

When moving from the bottom to the top of the classic data center topology how much is the increase in traffic?

Answer 29

200X

Question 30

What is the data center scale problem

Answer 30

every switch must store a forwarding table for every mac address

Question 31

What is one solution to the data center scale problem?

Answer 31

Group servers together as pods and assign a pseudo mac address to each pod

Question 32

What problem is faced when dealing with pseudo mac addresses?

Answer 32

Mapping pseudo mac to real mac, Intercepting ARP

Question 33

How is ARP handled by pseudo MAC?

Answer 33

The switch intercepts the query, sends it to an entity called the fabric manager which returns the pmac

Question 34

What is the objective of VL2

Answer 34

Abstraction of big L2 switches

Question 35

Goals of valiant load balacing

Answer 35

1 spread traffic, 2 Locate independance

Question 36

How does valiant load balancing work?

Answer 36

The traffic is forwarded to a random indirection point (switch) and then forwarded to the correct location

Question 37

Goals of Jellyfish

Answer 37

High throughput, Incremental expadability

Question 38

What is Jellyfish’s topology?

Answer 38

Random Regular Graph

Question 39

In Jellyfish where does the random selection come from?

Answer 39

From a regular graph

Question 40

What is a regular graph

Answer 40

Each node has same degree

Question 41

What makes up the graph

Answer 41

switches are nodes

Question 42

Jellyfish construction formula

Answer 42

i = number of switches, Ki = total ports, ri = ports used to connect to other switches, Ki - ri = ports connected to servers, N = number of racks, N(Ki - ri) servers are supported.

Question 43

Jellyfish construction algorithm

Answer 43

1. Pick a random switch pair with free ports that are not already neighbors, 2. Next join them with a link, 3. Repeat until no other links can be added

Question 44

What is the capacity increase of Jellyfish

Answer 44

25% more servers

Question 45

Higher capacity is achieved how?

Answer 45

Shorter paths

Question 46

Open Question related to Jellyfish

Answer 46

Topology Desigh: How close are random graphs to optimal, What about heterogeneous switches?

System Design: Cabling? Routing?