Midterm 3 Flashcards

1
Q

What are SDNs 2 defining characteristics?

A
  1. It separates the control plane from the data plane

2. It consolidates the control plane, so that a single software control program controls multiple data-plane elements

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is a “control plane”?

A

It decides how to handle the traffic; it computes the routing tables (routers have state called “routing tables”). Essentially, the logic that controls forwarding behavior e.g. routing protocols, configuration for network middleboxes, etc.

Typically written in a high level language like C or Python

Note: Today, control and data plane are on routers.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is a “data plane”?

A

It forwards traffic according to decisions that the control plane makes e.g. forwarding, switching

Typically programmable hardware controlled by control plane

Note: Today, control and data plane are on routers.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the “capsule model”?

A

Where the code to execute at the nodes was carried in-band in data packets

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is the “programmable routers/switch model”?

A

Where the code to execute at the nodes was established by out-of-band mechanisms

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What were the “technology pushes” that encouraged active networking?

A
  • reduction in cost of computing
  • making it conceivable to put more processing in the network
  • advances in programming languages such as Java that offered platform portability and code execution safety
  • virtual machine technology that protected the host machine
  • advances in rapid code compilation and formal methods
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What were the “use pulls” that encouraged active networking?

A
  • network service provider frustration with the timescales necessary to develop and deploy new network services (network ossification)
  • third party interest in value-added, fine-grained control to dynamically meet the needs of particular applications or network conditions
  • researcher desire for a platform that would support experimentation at scale
  • proliferation of middleboxes (including firewalls), proxies, and transcoders
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is Network Management?

A

Process of configuring a network to achieve a variety of tasks such as:

  • Load Balance across network
  • Security
  • Business Relationships that exist between configured network and neighbor networks
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Why is network configuration hard?

A
  • Defining correctness is hard
  • Interactions between protocols can lead to unpredictability; furthermore each AS is independently configured and interaction between policies can lead to unintended behavior
  • Operators make mistakes (device-level configuration, but SDN changes this)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What does SDN provide?

A
  1. Network-wide views of both Topology and Traffic
  2. Network-level objectives (Load Balance, Security, etc.)
  3. Direct control of data plane
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What should Routers do?

A
  • Forward Packets
  • Collect measurements

But not…
- Compute Routes (can be logically centralized)

SDN === “remove routing from routers”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What are the advantages of SDN (over conventional networks)?

A
  1. Coordination is easier
  2. Behavior of network is easier to evolve
  3. Reasoning is easier
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What are the applications of SDN?

A
  • Data Centers
  • Backbone networks
  • Enterprise networks
  • Internet Exchange Points (IXP)
  • Home Networks
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Name examples of Control Plane Operations

A

Job of Control Plane is to compute state that ultimately ends up in the data plane. Therefore,

  • computing a forwarding path that satisfies a high-level policy
  • computing shortest path routing trees
  • authenticating a user’s device based on a MAC address
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Why is separating data and control a good idea?

A
  1. Independent evolution and development thus software and hardware can evolve independently
  2. Control from high-level program allowing debugging and checking to be easier
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

How does control/data separation make managing data centers easier?

A
  1. Monitoring/control of routes from a central point

2. Migrating VMs without renumbering host addresses

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What is the Pox Algorithm for a switch?

A
  1. Update address/port table
  2. If packet is multicast, flood
  3. If no table entry, flood
  4. If src == dst, controller instructs switch to drop packet
  5. Install flow table entry
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What problems can arise from inconsistent writes of network state?

A
  • Forwarding loops
  • Security policy violations
  • A flood of traffic at the controller
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What is a Two-Phase Commit in terms of writing consistent policy?

A

Tag packet on ingress so that switches maintain copies of old and new policy; when all switches have received the new policy, incoming packets can be tagged with new policy version. After some time when no more packets are tagged with the old policy, it can be removed.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

What is network virtualization?

A

Abstraction of physical network where multiple logical networks can be run on the same underlying shared physical substrate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Why Virtual Networking?

A

Ossificiation of internet architecture aka because the internet protocol was so evasive, it made it very difficult to make fundamental changes to the way the underlying internet architecture operated.

Rather than replace existing network architectures, network virtualization was intended to allow for easier evolution.

“Network virtualization enables evolution by letting multiple architectures exist in parallel.”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What is the motivation for virtual networking?

A
  • Facilitating research/evolution by allowing coexistence

- Adjusting resources to demand

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

What are some benefits of network virtualization?

A

+ Rapid innovation (software speed)
+ New forms of network control
+ (Potentially) simpler programming

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

What is the difference between SDN and Network Virtualization?

A

Network Virtualization is arguably one of the first killer applications of SDN and in some sense, SDN is a tool for implementing Network Virtualization.

SDN separates data and control plane. Network Virtualization separates logical and physical layer. SDN is used to simplify many aspects of Network Virtualization, but doesn’t abstract details of underlying network.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

What are the design goals of Network Virtualization?

A
  • Flexible
  • Manageable
  • Scalable
  • Secure
  • Programmable
  • Heterogeneous (able to support different technologies)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What are the 2 different ways to compose policies?

A

Parallel: Perform both operations simultaneously (e..g counting and forwarding)

Sequential: Perform one operation, then the next (e.g. firewall, then switch)

27
Q

What is Traffic Engineering?

A

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

28
Q

What is Intradomain Traffic Engineering?

A
  • Routes flood information to learn topology
  • Operator configures link weights (inversely proportional to capacity, proportional to propagation delay, network-wide optimization based on traffic)
29
Q

What is Interdomain Traffic Engineering?

A
  • Done by reconfiguring BGP
  • Alleviating congestion on edge links
  • Using new/upgraded edge links
  • Changing end-to-end path
30
Q

What are the 3 steps of Intradomain Traffic Engineering?

A
  1. Measure network to figure out current traffic
  2. Model of how configuration affects underlying paths
  3. Exert control by reconfiguring network
31
Q

What is the goal of Jellyfish?

A
  1. High throughput (big data analytics)

2. Incremental expandability (easy replacement of servers)

32
Q

What are some cons of using Layer 3 forwarding?

A
  • The process of adding a new switch requires manual administrator configuring and oversight, which is prone to error
  • Improperly synchronized state between system components, such as a DHCP server and a configured switch subnet identifier can lead to unreachable hosts and are difficult to diagnose
  • Growing importance of end host virtualization makes Layer 3 solutions less desirable
33
Q

What are pros of using Layer 2 forwarding?

A

Forwarding is performed based on flat MAC addresses

  • Standard Ethernet bridging does not scale to networks with tens of thousands of hosts because of the need to support broadcast across the entire fabric
  • Single forwarding spanning tree limits performance in topologies that consist of multiple available equal cost paths
34
Q

What is the middle ground between Layer 2 and Layer 3 forwarding?

A

Employing VLANs to allow a single logical Layer 2 fabric across multiple switch boundaries

35
Q

What are some cons from using VLANs?

A
  • Requires bandwidth resources to be explicitly assigned to each VLAN at each participating switch, limiting flexibility for dynamically changing communication patterns
  • Each switch must maintain state for all hosts in each VLAN that they participate in, limiting scalability
  • Single forwarding spanning tree, limiting performance
36
Q

What are the means for internet attacks?

A
  • Routing (BGP)

- Naming (DNS) via reflection (DDoS) or phishing

37
Q

What makes the internet’s design fundamentally insecure?

A
  • On by default design
  • Attacks look like normal traffic
  • Internet is federated
38
Q

What are the components of internet security?

A

Availability: ability to use a resource
Confidentiality: concealing information
Authenticity: assures origin of information
Integrity: prevent unauthorized changes

39
Q

What is the difference between a security threat and an attack?

A

A threat is a potential violation whereas an attack is action that results in a violation of availability, confidentiality, authenticity, or integrity

40
Q

A denial of service is an attack of which property of internet security?

A

Availability

41
Q

Why is DNS vulnerable?

A
  • Queries over UDP
  • No authentication for query responses

Essentially there is no authentication. The resolvers trust the responses that are received after the query regardless of where they come from. These responses can be forged by an attacker.

Responses can also contain info unrelated to query. Queries are typically connectionless and therefore resolver doesn’t have a way to map a response other than the query ID, which can be forged by an attacker.

42
Q

What are defenses against DNS Cache Poisoning?

A
  1. Query ID (can be guessed)
  2. Randomize ID (makes ID tougher to guess, but ID is only 16 bits - birthday paradox)
  3. Source port randomization
  4. Ox20 encoding (combo of upper and lowercase letters, so guessing is much more difficult)

Attacker must also win “race”

43
Q

What is the DNSSEC Protocol?

A

DNSSEC adds authentication to DNS responses by adding signatures to responses returned for each DNS reply

44
Q

What is Bulletproof Hosting?

A

Service providers that willingly host a malicious autonomous system where cyber-criminals host their services and protects illicit activities. The name refers to their reluctance to address repeated abuse complaints regarding their customers and the illegal services they run.

45
Q

How can ASwatch help network operators and security practitioners?

A
  1. Prioritize traffic - knowing what ASes have suspicious control-plane behavior may help admins appropriately handle traffic
  2. Peering decisions - upstream providers could use AS reputation scores as an additional source of info to make peering decisions
  3. Prioritize investigations - law enforcement and security practitioners may prioritize their investigations and start early mornitoring on low reputation ASes
  4. Complement data-plane based systems - ASes that exhibit malicious behavior from both can be prioritize first
  5. Strengthen existing defenses - reputation could be used as input to other network defenses to improve their detection accuracy e.g. spam filters, botnet detection systems
46
Q

What is a virus?

A

Infection of an existing program that results in a modification of the original program’s behavior.

Viruses spread manually and require user action.

47
Q

What is a worm?

A

Code that propagates/replicate itself across the network. Usually spread by exploiting flaws in existing programs/services.

Worms spread automatically.

48
Q

What are the 4 types of viruses?

A
  1. Parasitic - infects executable files
  2. Memory Resident - infects running programs
  3. Boot-sector - spreads when system is booted
  4. Polymorphic - encrypt part of virus program using randomly generated key
49
Q

What is a Psuedo Mac Address (PMAC)?

A

PortLand further assigns internal Pseudo MAC

(PMAC) addresses to all end hosts to encode their position in the topology.

50
Q

What is a Fabric Manager used for?

A

PortLand employs a logically centralized fabric manager that maintains soft state about network configuration information such as topology. The fabric manager is a user process running on a dedicated machine responsible for assisting with ARP resolution, fault tolerance, and multicast

51
Q

What are the 4 parts of a PMAC?

A

pod
position
port
vmid

52
Q

How does PMAC reduce forwarding table sizes?

A

Only routing table on the AMAC machine are the PMACS that belong to it (and vice versa).

53
Q

What is Active Networking?

A

A programming interface (or network API) that exposes resources (e.g. processing, storage, and packet queues) on individual network nodes.

Supports construction of custom functionality to apply to a subset of packets.

54
Q

What are the 3 stages in history of SDN?

A

Active Networks (1990s-2000s): introduced programmable functions in the network to enable greater innovation

Control and Data Plane Separation (2001-2007): developed open interfaces between the control and data planes

OpenFlow API and network operating systems (from 2007-2010): first instance of widespread adoption of open interface, developed ways to make control-data pane separation scalable and practical

55
Q

What is PortLand?

A

a scalable, fault tolerant layer 2 routing and forwarding protocol for data center environments. PortLand holds promise for supporting a “plug-and-play” large-scale, data center network. The goal of PortLand is to deliver scalable layer 2 routing, forwarding, and addressing for data center network environments.

56
Q

Why is BGP susceptible to attacks?

A

BGP allows any AS to advertise an IP prefix to a neighboring AS, and that AS will typically just believe that route advertisement and advertise it to the rest of the internet. These events that occur where an AS advertises a prefix that it does not own are called route highjacks.

57
Q

Why is the Internet fundamentally insecure?

A

The internet was designed for simplicity, and as a result security was not a primary consideration when the internet was originally designed it’s on by default. In other words, when a host is connected to the internet, it is by default reachable by any other host that has a public IP address. This means that if one has an insecure host, that host is effectively wide open to attack by other hosts on the internet. the internet is run by tens of thousands of independently run networks, it can be very difficult to coordinate a defense against an attack because each of these networks is run by different network operators, sometimes in completely different countries.

58
Q

Why does the 0x20 encoding make DNS more secure?

A

The 0x20 bit encoding adds additional entropy to the queries that a DNS resolver sends by tweaking the capitilization on a DNS name in such a way that only the resolver and the authoritative name server know the particular sequence of upper and lower case letters in the reply.

59
Q

What are the goals of valiant load balancing?

A
  1. Spread traffic evenly

2. Traffic flows are balance independently

60
Q

A route hijack is an attack on what?

A

Origin Authentication

61
Q

What are the problems with spam?

A
  1. Filters - separating good form bad
  2. Storage
  3. Security Problem - phishing
62
Q

How to differentiate spam from “ham”?

A
  1. content based
  2. IP address of sender
  3. Behavioral Features e.g. filter based on how message is sent
63
Q

What are defenses against Denial of Service Attacks?

A
  1. Ingress filtering
  2. uRPF checks
  3. Syn Cookies (TCP)
64
Q

Goals of Interdomain Traffic Engineering

A
  1. Predictability
  2. Limit influence of neighbors
  3. Reduce overload of routing changes