Tutorial 4: 31st October 2019 Flashcards

Datacentre networking

1
Q

What are elephant flows?

A

In computer networking, an elephant flow is an extremely large (in total bytes) continuous flow set up by a TCP (or other protocol) flow measured over a network link. Elephant flows, though not numerous, can occupy a disproportionate share of the total bandwidth over a period of time.

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

What are mouse flows?

A

In computer networking, a mouse flow is a short (in total bytes) flow set up by a TCP (or other protocol) flow measured over a network link.

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

How do elephant flows affect mouse flows?

A

Elephant flows cause delays in mouse flows, reducing their performance.

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

What is an autonomous system?

A

A network or a collection of networks that are all managed and supervised by a single entity or organization. They use a single routing mechanism to connect to the Internet.

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

What is BGP?

A

BGP = Border Gateway Protocol = A standardized exterior gateway protocol designed to exchange routing and reachability information among autonomous systems (AS) on the Internet.

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

How do the AS and ISP models compare?

A

Lack of homogeneity in ASes may make it harder to reach everyone and shard the Internet. Easy for ISPs to connect users into IXPs.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is a Point of Presence?

A

A point of presence is a chosen interface point betwee and ISP and home LANs.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How do the Point-of-Presence and ISP models compare?

A

ISPs are not-for-profit, POPs are. POP could lead to sharding and lack of neutrality.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the main difference between networks in datacentres and the wider Internet?

A

You can control all parts of a datacentre network and have total deployment of protocols. You have homogeneity and symmetry.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is FOG computing?

A

Using edge devices to carry out a substantial amount of computation, storage, communication locally and routed over the internet backbone.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

How does having few large datacentres compare to having many small datacentres?

A

Lower partition tolerance, higher latency, and more difficult to design, manage, and build individually. Probably lower cost and complexity overall though.

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

What are the performance priorities in cloud applications?

A

availability and low latency

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How can you determine what a user is doing from their outgoing HTTP/HTTPS requests?

A

Look at IP of first request, then IPs after (will be libraries, static resources, etc). You can guess with a high probability what the URL originally requested was.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

How could a hybrid between few large datacentres and many small ones work?

A

Use large ones as authoritative sources and smaller ones ~ as caches. Smaller ones serve simple content like HTML, image, and CSS (static) files and DB queries to larger ones? Or even just HTML in smaller ones so something is loading and rest in large ones.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is Akami?

A

An American content delivery network (CDN), cybersecurity, and cloud service provider. Akamai’s content delivery network is one of the world’s largest distributed computing platforms, responsible for serving between 15% and 30% of all web traffic.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Why can no single network topology present a final optimal solution to datacentre networking?

A

Independent groups will implement them differently and they will be implemented differently on different hardware, leading to different results. Also, all have pros and cons in tradeoffs: use different ones in different situations.

17
Q

How does TCP implement congestion signalling?

A

With ECN bits: Explicit Congestion Notification. The bit in TCP headers set to 1 when congestion detected by the receiver, leading to a halving of bandwidth each window where an ECN bit is set to 1.

18
Q

What are the limitations of ECN bits?

A

Can’t estimate or measure the magnitude of congestion, nor locate it.