优化多CDN：揭秘Netflix视频流媒体策略与带宽提升

Netflix

CDN

需积分: 10 5 浏览量更新于2024-09-10 收藏 941KB PDF 举报

身份认证购VIP最低享 7 折!

30元优惠券

资源详情

资源推荐

Unreeling Netﬂix: Understanding and Improving

Multi-CDN Movie Delivery

Vijay Kumar Adhikari

∗

, Yang Guo

†

, Fang Hao

†

, Matteo Varvello

†

, Volker Hilt

†

, Moritz Steiner

†

and Zhi-Li Zhang

∗

University of Minnesota,

†

Bell-Labs/Alcatel-Lucent

Abstract—Netﬂix is the leading provider of on-demand Internet

video streaming in the US and Canada, accounting for 29.7%

of the peak downstream trafﬁc in US. Understanding the Netﬂix

architecture and its performance can shed light on how to best

optimize its design as well as on the design of similar on-demand

streaming services. In this paper, we perform a measurement

study of Netﬂix to uncover its architecture and service strategy.

We ﬁnd that Netﬂix employs a blend of data centers and Content

Delivery Networks (CDNs) for content distribution. We also

perform active measurements of the three CDNs employed by

Netﬂix to quantify the video delivery bandwidth available to

users across the US. Finally, as improvements to Netﬂix’s current

CDN assignment strategy, we propose a measurement-based

adaptive CDN selection strategy and a multiple-CDN-based video

delivery strategy, and demonstrate their potentials in signiﬁcantly

increasing user’s average bandwidth.

I. INTRODUCTION

Netﬂix is the leading subscription service provider for

online movies and TV shows. Netﬂix attracts more than 23

million subscribers in the United States and Canada, and can

stream out HD (High Deﬁnition) quality video with average

bitrate reaching 3.6 Mbps. In fact, Netﬂix is the single largest

source of Internet trafﬁc in the US, consuming 29.7% of peak

downstream trafﬁc [1]. Its design and trafﬁc management deci-

sions have a substantial impact on the network infrastructure.

Designing such a large scale, fast growing video streaming

platform with high availability and scalability is technically

challenging. The majority of functions used to be hosted in

Netﬂix’s own data center. Recently, Netﬂix has resorted to

the use of cloud services [2], Content Distribution Networks

(CDNs), and other public computing services. Amazon AWS

cloud replace in-house IT, and Amazon SimpleDB, S3 and

Cassandra are used for ﬁle storage [2]. Video streaming is

served out of multiple CDNs, and UltraDNS, a public DNS

service, is used as its authoritative DNS servers. Microsoft

Silverlight [3] is employed as the video playback platform

for Netﬂix desktop users. The end result is amazing: Netﬂix

manages to build its Internet video delivery service with little

infrastructure of its own!

In this paper we provide a detailed analysis of the Netﬂix

architecture, which is designed to serve massive amounts

of content by combining multiple third party services. This

This study was done while Vijay Kumar Adhikari was a summer intern

at Bell Labs, Alcatel-Lucent. It was supported in part by the NSF grants

08-31734, CNS-0905037, CNS-1017092 and CNS-1017647.

architecture can be considered as a possible blue print for a

scalable, infrastructure-less content provider. We discuss the

interaction between the components of this design including

multiple CDNs and HTTP adaptive streaming, and analyze

the algorithms used by Netﬂix that provide the glue to piece

together the overall system. We also shed light on the implica-

tions the Netﬂix design decisions have on CDNs, the network

and the end user experience both to understand its performance

and to improve its design. In addition, based on our measure-

ment results, we suggest new video delivery strategies that

can further improve user experience by effectively utilizing

multiple CDNs.

Despite the popularity of Netﬂix, surprisingly there have

been very few studies looking into its streaming service

platform. The authors of [4] investigate the Netﬂix secu-

rity framework, while the authors of [5] focus on the rate-

adaptation mechanisms employed by Silverlight player and

experimentally evaluated the Netﬂix players. To the best of

our knowledge, we are the ﬁrst to take a systematic look into

the architecture of the Netﬂix video streaming together with

an extensive measurement study of three CDNs it employs.

The main contributions of this paper can be summarized as

follows:

• We dissect the basic architecture of the Netﬂix video

streaming platform by monitoring the communications

between the Netﬂix player and various components of

the Netﬂix platform. We collect a large number of Netﬂix

video streaming manifest ﬁles to analyze how geographic

locations, client capabilities, and content type affect the

streaming parameters used by Netﬂix, such as content

formats, video quality levels, CDN ranking, and so forth.

• We analyze how Netﬂix makes use of multiple CDNs un-

der changing bandwidth conditions. We ﬁnd that Netﬂix

players stay attached to a ﬁxed CDN even when the other

CDNs can offer better video quality.

• We perform an extensive bandwidth measurement study

of the three CDNs used by Netﬂix. The results show that

there is signiﬁcant variation in CDN performance across

time and location.

• Finally, we explore alternative strategies for improving

video delivery performance by using multiple CDNs. Our

study shows that selecting the best serving CDN based

on a small number of measurements at the beginning of

each video session can deliver more than 12% bandwidth

下载后可阅读完整内容，剩余8页未读，立即下载

qq_16095421

粉丝: 0
资源: 1

优化多CDN：揭秘Netflix视频流媒体策略与带宽提升

spring-cloud-netflix-hystrix应用

IMDb & Rotten Tomatoes for Netflix:trade_mark:-crx插件

springcloud依赖

Cannot resolve org.springframework.cloud:spring-cloud-starter-netflix-hystrix:2.0.1.RELEASE 启动项目

'dependencies.dependency.version' for org.springframework.cloud:spring-cloud-starter-netflix-eureka-server:jar is missing.

org.springframework.cloud:spring-cloud-starter-netflix-eureka-server的maven依赖

Dependency 'org.springframework.cloud:spring-cloud-starter-netflix-hystrix-dashboard:2.2.10.RELEASE' not found

eureka配置https

springboot gateway 如何配置负载均衡策略

ork.cloud:spring-cloud-starter-openfeign:jar:

写一个java eureka客户端

springcloud依赖文件对应关系

spring-cloud-starter-netflix-zuul和spring-cloud-starter-zuul区别

Could not transfer artifact org.springframework.cloud:spring-cloud-starter-openfeign:pom:2.2.1.RELEASE

spring-cloud-starter-netflix-eureka-client与spring-cloud-starter-eureka的区别

Java种和spring-boot-start-parent2.4.13匹配的eureka版本

Dependency 'org.springframework.cloud:spring-cloud-starter-netflix-hystrix:' not found

no main manifest attribute, in spring-cloud-starter-netflix-eureka-server-2.2.6.RELEASE.jar

使用python爬取https://www.kaggle.com/datasets/ariyoomotade/netflix-data-cleaning-analysis-and-visualization上的数据集，并保存在本地文件中，给出详细代码

最新资源