Scribd
Upload
52 views17 pages

L1 Intro

This document provides an introduction to distributed systems. It defines distributed systems as involving multiple connected computers working together over a network. Distributed systems are used to address issues like failure, limited resources, and physical location. Examples of distributed systems include web search, e-commerce, file syncing, social networks, music streaming, and more. The goal of distributed systems is to provide higher-level abstractions that hide complexity while delivering scalability, reliability, and consistency guarantees.

Uploaded by

abbasalizaidi07
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
52 views17 pages

L1 Intro

This document provides an introduction to distributed systems. It defines distributed systems as involving multiple connected computers working together over a network. Distributed systems are used to address issues like failure, limited resources, and physical location. Examples of distributed systems include web search, e-commerce, file syncing, social networks, music streaming, and more. The goal of distributed systems is to provide higher-level abstractions that hide complexity while delivering scalability, reliability, and consistency guarantees.

Uploaded by

abbasalizaidi07
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Distributed Systems Intro

COS 418/518: Distributed Systems


Lecture 1

Wyatt Lloyd
Distributed Systems, What?

1) Multiple computers
2) Connected by a network
3) Doing something together
Distributed Systems, Why?
• Or, why not 1 computer to rule them all?

• Failure

• Limited computation/storage/…

• Physical location
Distributed Systems, Where?
• Web Search (e.g., Google, Bing)
• Shopping (e.g., Amazon, Walmart)
• File Sync (e.g., Dropbox, iCloud)
• Social Networks (e.g., Facebook, Twitter)
• Music (e.g., Spotify, Apple Music)
• Ride Sharing (e.g., Uber, Lyft)
• Video (e.g., Youtube, Netflix)
• Online gaming (e.g., Fortnite, DOTA2)
• …
“The Cloud” is not
amorphous

5
Microsoft 6
Google

7
Facebook 8
Facebook 9
100,000s of physical servers
10s MW energy consumption

Facebook Prineville:
$250M physical infra, $1B IT infra
Distributed Systems Goal
• Service with higher-level abstractions/interface
• e.g., file system, database, key-value store, programming
model, …

• Hide complexity
• Scalable (scale-out)
• Reliable (fault-tolerant)
• Well-defined semantics (consistent)

• Do “heavy lifting” so app developer doesn’t need to


Scalable Systems in this Class
• Scale computation across many machines
• MapReduce, Streaming Video Engine

• Scale storage across many machines


• Dynamo, COPS, Spanner
Fault Tolerant Systems in this Class
• Retry on another machine
• MapReduce, Streaming Video Engine

• Maintain replicas on multiple machines


• Primary-backup replication
• Paxos
• RAFT
• Bayou
• Dynamo, COPS, Spanner
Range of Abstractions and Guarantees
• Eventual Consistency
• Dynamo

• Causal Consistency
• Bayou, COPS

• Linearizability
• Paxos, RAFT, Primary-backup replication

• Strict Serializability
• 2PL, Spanner
Learning Objectives
• Reasoning about concurrency
• Reasoning about failure
• Reasoning about performance

• Building systems that correctly handle concurrency and failure

• Knowing specific system designs and design components


Conclusion
• Distributed Systems
• Multiple machines doing something together
• Pretty much everywhere and everything computing now

• “Systems”
• Hide complexity and do the heavy lifting (i.e., interesting!)
• Scalability, fault tolerance, guarantees

You might also like