Distributed Systems

?
  • Created by: smrc
  • Created on: 02-05-19 15:14
State the Definition of a Distributed System
A distributed system is: a collection of independent computers that appears to its users as a single coherent system.
1 of 22
How do components in a distributed system communicate?
Components (located at networked computers) communicate with each other and coordinate their actions via message passing.
2 of 22
What is the main motivation for using distributed systems? Give an example of an application.
Purpose: To share resources. Examples: Printers, files, web pages, database records, hardware devices
3 of 22
List 3 characteristics of a Distributed System.
1. No global clock 2. Concurrency 3. Independent failures (network or computer).
4 of 22
Explain what is meant by NO GLOBAL CLOCKs in a Distributed System.
Nodes coordinate by exchanging messages. Because of this there are limits to how accurately their clocks can be synchronised.
5 of 22
Explain what is meant by Concurrency in a Distributed System.
The issue of coordinating the execution of programs that share resources concurrently.
6 of 22
What is the issue with INDEPENDENT FAILURES (network or computer) in a distributed system?
Other components in the system may not be aware of the failure/s.
7 of 22
List the important ways in which distributed computing systems are inherently different from non-distributed systems, and briefly explain their significance.
1. LATENCY (network speed) 2. MEMORY ACCESS 3. PARTIAL FAILURE 4. GUARANTEED CONCURRENCY (with latency) 5. INDETERMINACY
8 of 22
Explain how the network speed (LATENCY) differs in a DS.
The time taken to transmit a packet in a DS is greater than a non-DS.
9 of 22
Explain how the MEMORY ACCESS differs in a DS.
There is no shared memory in a DS.
10 of 22
Explain how the failures (PARTIAL FAILURE) differs in a DS.
Partial failures can occur in DSs. A remote failure occurring does not mean that a local failure has occurred
11 of 22
Explain the concept of GUARANTEED CONCURRENCY (with latency) in DSs.
Guaranteed concurrency, combined with latency means that events are not necessarily received in the same order as they are generated. It is impossible to determine the order in which remote events take place.
12 of 22
Explain the concept of INDETERMINACY in DSs.
Local system not in control of whole system in DSs. No indication given if a partial failure occurs. Difficult to determine cause/origin of failures.
13 of 22
Explain why Peer-to-Peer (P2P) systems are better than Centralised systems when sharing data and resources on a very large scale.
Traditional client-server systems use centralised designs. In these systems the scale of the services is limited by their hardware capacity and network connectivity. P2P middleware has capacity to share computing resources, storage + data globally.
14 of 22
What is the purpose of P2P systems?
P2P systems aim to support distributed services and applications using data and computing resources available on the internet.
15 of 22
What are 4 characteristics of nodes in a P2P system?
All the nodes in the system: 1. are resource providers as well as consumers. 2. are equal in terms of functional capacities and responsibilities. 3. Operate independently from a centrally administered system. 4. are provided with degree of anonymity
16 of 22
Give 3 examples of real world applications of P2P systems.
Napster, Gnutella, BitTorrent
17 of 22
Briefly explain the P2P system: Napster
Napster: The 1st P2P file sharing system that allowed nodes to share files based on centralised indexes.
18 of 22
Briefly explain the P2P system: Gnutella
Gnutella: 2nd gen P2P file sharing app that avoided the need for centralised indexes.
19 of 22
Briefly explain the P2P system: BitTorrent
P2P sharing for very large files. Breaks files into chunks that at any given time are distributed over P2P nodes. A node wanting a particular file has to assemble it from the chunks held by other nodes.
20 of 22
Briefly explain the advantages of P2P systems.
1. CAN EXPLOIT UNUSED RESOURCES 2. GOOD LOAD BALANCING on network links and host computing resources, can scale well 3. SELF ORGANISING PROPERTIES means lower support costs because the costs are independent of the number of clients and hosts (nodes)
21 of 22
Briefly explain the disadvantages of P2P systems.
1. Use for storage of mutable data costly compared with a trusted centralised service. 2. No strong guarantee of anonymity. 3. Issues of maintaining coherent state and consensus my be important (transactions)
22 of 22

Other cards in this set

Card 2

Front

How do components in a distributed system communicate?

Back

Components (located at networked computers) communicate with each other and coordinate their actions via message passing.

Card 3

Front

What is the main motivation for using distributed systems? Give an example of an application.

Back

Preview of the front of card 3

Card 4

Front

List 3 characteristics of a Distributed System.

Back

Preview of the front of card 4

Card 5

Front

Explain what is meant by NO GLOBAL CLOCKs in a Distributed System.

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Computing resources:

See all Computing resources »See all Emerging Technologies resources »