Classification of load balancing in the internet
Autor(a) principal: | |
---|---|
Data de Publicação: | 2019 |
Tipo de documento: | Dissertação |
Idioma: | por |
Título da fonte: | Repositório Institucional da UFMG |
Texto Completo: | http://hdl.handle.net/1843/34630 |
Resumo: | A router may perform load balancing and distribute traffic across multiple routes that have the same cost. Load balancing improves available bandwidth, robustness to failures, and performance. Routers that perform load balancing (referred to as load balancers) compute the link a packet should be forwarded to as a function of the packet’s flow identifier, a subset of fields in the packet’s headers (e.g., IP addresses and port numbers). Network operators and researchers rely on measurement tools to identify and characterize load balancing. However, recent advances in programmable data planes, software defined networks, and even the adoption of IPv6, support novel, more complex load balancing strategies. These strategies allow the definition of flow identifiers that existing measurement tools are incompatible with. In this work, we introduce the Multipath Classification Algorithm (MCA). We generalize the network formalism used to describe load balancing and extend existing techniques to consider that load balancers may use arbitrary combinations of packet header fields for load balancing. MCA detects load balancers that existing tools cannot, regardless of the bits load balancers consider in flow identifiers. Furthermore, MCA classifies the behavior of load balancers and their impact on application traffic. We propose optimizations that reduce the classification cost by 11% and the overall cost by 6%, without loss of accuracy. Our evaluation shows that the process of classifying load balancers entails a cost similar to the cost of the detection process, demonstrating MCA is a practical tool. Finally, we use MCA to collect a representative dataset of route measurements to characterize load balancing in the Internet. Our results show that load balancing is more prevalent and load balancing strategies are more mature than previous characterizations have found. |
id |
UFMG_d45a67a109a76f753b00b0373ba2e0ca |
---|---|
oai_identifier_str |
oai:repositorio.ufmg.br:1843/34629 |
network_acronym_str |
UFMG |
network_name_str |
Repositório Institucional da UFMG |
repository_id_str |
|
spelling |
Italo Fernando Scota Cunhahttp://lattes.cnpq.br/7973706384467274Renata Cruz TeixeiraLuiz Felipe Menezes VieiraAna Paula Couto da SilvaAntonio Alfredo Ferreira Loureirohttp://lattes.cnpq.br/6020676623308883Rafael Luis Caldas Almeida2021-01-05T18:54:35Z2021-01-05T18:54:35Z2019-07-01http://hdl.handle.net/1843/34630A router may perform load balancing and distribute traffic across multiple routes that have the same cost. Load balancing improves available bandwidth, robustness to failures, and performance. Routers that perform load balancing (referred to as load balancers) compute the link a packet should be forwarded to as a function of the packet’s flow identifier, a subset of fields in the packet’s headers (e.g., IP addresses and port numbers). Network operators and researchers rely on measurement tools to identify and characterize load balancing. However, recent advances in programmable data planes, software defined networks, and even the adoption of IPv6, support novel, more complex load balancing strategies. These strategies allow the definition of flow identifiers that existing measurement tools are incompatible with. In this work, we introduce the Multipath Classification Algorithm (MCA). We generalize the network formalism used to describe load balancing and extend existing techniques to consider that load balancers may use arbitrary combinations of packet header fields for load balancing. MCA detects load balancers that existing tools cannot, regardless of the bits load balancers consider in flow identifiers. Furthermore, MCA classifies the behavior of load balancers and their impact on application traffic. We propose optimizations that reduce the classification cost by 11% and the overall cost by 6%, without loss of accuracy. Our evaluation shows that the process of classifying load balancers entails a cost similar to the cost of the detection process, demonstrating MCA is a practical tool. Finally, we use MCA to collect a representative dataset of route measurements to characterize load balancing in the Internet. Our results show that load balancing is more prevalent and load balancing strategies are more mature than previous characterizations have found.Um roteador pode realizar balanceamento de carga e distribuir tráfego entre múltiplas rotas que têm o mesmo custo. Balanceamento de carga melhora a banda disponível, robustez a falhas e desempenho. Roteadores que fazem balanceamento de carga (chamados de balanceadores de carga) calculam qual enlace cada pacote deve ser encaminhado em função do identificador de fluxo, um subconjunto de campos nos cabeçalhos do pacote (e.g., endereços IP e números de porto). Operadores de rede e pesquisadores dependem de ferramentas de medição que identifiquem balanceamento de carga e caracterizem seu comportamento. No entanto, avanços recentes em planos de dados programáveis, redes definidas por software e até mesmo a adoção de IPv6 suportam novas e mais complexas estratégias de balanceamento de carga, permitindo a definição de identificadores de fluxo incompatíveis com ferramentas existentes. Neste trabalho, introduzimos o Multipath Classification Algorithm (MCA). Generalizamos o formalismo de rede utilizado para descrever balanceamento de carga e estendemos técnicas existentes para o cenário onde balanceadores de carga podem usar identificadores de fluxo compostos por combinações arbitrárias de bits nos cabeçalhos dos pacotes. O MCA detecta balanceadores de carga que técnicas existentes são incapazes de detectar, independente de quais bits compõem os identificadores de fluxo. Além disso, o MCA permite classificar o comportamento de cada balanceador de carga e seu impacto sobre o tráfego de aplicações. Para limitar o custo de medições usando MCA, propomos otimizações que reduzem o custo da classificação em 11% e o custo global em 6%, sem perda de acurácia. Nossa avaliação mostra que o processo de classificação acarreta um custo semelhante ao custo do processo de detecção, demonstrando a utilidade prática do MCA. Por fim, utilizamos o MCA para coletar um conjunto de dados representativo de rotas na Internet para caracterizar o balanceamento de carga na Internet. Nossos resultados mostram que o balanceamento de carga na Internet hoje é mais prevalente e mais moderno em relação a caracterizações anteriores.porUniversidade Federal de Minas GeraisPrograma de Pós-Graduação em Ciência da ComputaçãoUFMGBrasilICX - DEPARTAMENTO DE CIÊNCIA DA COMPUTAÇÃOComputação – Teses.Redes de computadores - Teses.Internet – Protocolos – Teses.IPv6 – TesesInternetRoutingProtocolsComputer networkingClassification of load balancing in the internetinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisinfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da UFMGinstname:Universidade Federal de Minas Gerais (UFMG)instacron:UFMGORIGINALmda.pdfmda.pdfapplication/pdf996695https://repositorio.ufmg.br/bitstream/1843/34629/2/mda.pdff8209fe4a75e797c5b4f282a18669c66MD52LICENSElicense.txtlicense.txttext/plain; charset=utf-82119https://repositorio.ufmg.br/bitstream/1843/34629/3/license.txt34badce4be7e31e3adb4575ae96af679MD531843/346292021-01-05 15:54:35.142oai:repositorio.ufmg.br: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Repositório de PublicaçõesPUBhttps://repositorio.ufmg.br/oaiopendoar:2021-01-05T18:54:35Repositório Institucional da UFMG - Universidade Federal de Minas Gerais (UFMG)false |
dc.title.pt_BR.fl_str_mv |
Classification of load balancing in the internet |
title |
Classification of load balancing in the internet |
spellingShingle |
Classification of load balancing in the internet Rafael Luis Caldas Almeida Internet Routing Protocols Computer networking Computação – Teses. Redes de computadores - Teses. Internet – Protocolos – Teses. IPv6 – Teses |
title_short |
Classification of load balancing in the internet |
title_full |
Classification of load balancing in the internet |
title_fullStr |
Classification of load balancing in the internet |
title_full_unstemmed |
Classification of load balancing in the internet |
title_sort |
Classification of load balancing in the internet |
author |
Rafael Luis Caldas Almeida |
author_facet |
Rafael Luis Caldas Almeida |
author_role |
author |
dc.contributor.advisor1.fl_str_mv |
Italo Fernando Scota Cunha |
dc.contributor.advisor1Lattes.fl_str_mv |
http://lattes.cnpq.br/7973706384467274 |
dc.contributor.referee1.fl_str_mv |
Renata Cruz Teixeira |
dc.contributor.referee2.fl_str_mv |
Luiz Felipe Menezes Vieira |
dc.contributor.referee3.fl_str_mv |
Ana Paula Couto da Silva |
dc.contributor.referee4.fl_str_mv |
Antonio Alfredo Ferreira Loureiro |
dc.contributor.authorLattes.fl_str_mv |
http://lattes.cnpq.br/6020676623308883 |
dc.contributor.author.fl_str_mv |
Rafael Luis Caldas Almeida |
contributor_str_mv |
Italo Fernando Scota Cunha Renata Cruz Teixeira Luiz Felipe Menezes Vieira Ana Paula Couto da Silva Antonio Alfredo Ferreira Loureiro |
dc.subject.por.fl_str_mv |
Internet Routing Protocols Computer networking |
topic |
Internet Routing Protocols Computer networking Computação – Teses. Redes de computadores - Teses. Internet – Protocolos – Teses. IPv6 – Teses |
dc.subject.other.pt_BR.fl_str_mv |
Computação – Teses. Redes de computadores - Teses. Internet – Protocolos – Teses. IPv6 – Teses |
description |
A router may perform load balancing and distribute traffic across multiple routes that have the same cost. Load balancing improves available bandwidth, robustness to failures, and performance. Routers that perform load balancing (referred to as load balancers) compute the link a packet should be forwarded to as a function of the packet’s flow identifier, a subset of fields in the packet’s headers (e.g., IP addresses and port numbers). Network operators and researchers rely on measurement tools to identify and characterize load balancing. However, recent advances in programmable data planes, software defined networks, and even the adoption of IPv6, support novel, more complex load balancing strategies. These strategies allow the definition of flow identifiers that existing measurement tools are incompatible with. In this work, we introduce the Multipath Classification Algorithm (MCA). We generalize the network formalism used to describe load balancing and extend existing techniques to consider that load balancers may use arbitrary combinations of packet header fields for load balancing. MCA detects load balancers that existing tools cannot, regardless of the bits load balancers consider in flow identifiers. Furthermore, MCA classifies the behavior of load balancers and their impact on application traffic. We propose optimizations that reduce the classification cost by 11% and the overall cost by 6%, without loss of accuracy. Our evaluation shows that the process of classifying load balancers entails a cost similar to the cost of the detection process, demonstrating MCA is a practical tool. Finally, we use MCA to collect a representative dataset of route measurements to characterize load balancing in the Internet. Our results show that load balancing is more prevalent and load balancing strategies are more mature than previous characterizations have found. |
publishDate |
2019 |
dc.date.issued.fl_str_mv |
2019-07-01 |
dc.date.accessioned.fl_str_mv |
2021-01-05T18:54:35Z |
dc.date.available.fl_str_mv |
2021-01-05T18:54:35Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/1843/34630 |
url |
http://hdl.handle.net/1843/34630 |
dc.language.iso.fl_str_mv |
por |
language |
por |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Universidade Federal de Minas Gerais |
dc.publisher.program.fl_str_mv |
Programa de Pós-Graduação em Ciência da Computação |
dc.publisher.initials.fl_str_mv |
UFMG |
dc.publisher.country.fl_str_mv |
Brasil |
dc.publisher.department.fl_str_mv |
ICX - DEPARTAMENTO DE CIÊNCIA DA COMPUTAÇÃO |
publisher.none.fl_str_mv |
Universidade Federal de Minas Gerais |
dc.source.none.fl_str_mv |
reponame:Repositório Institucional da UFMG instname:Universidade Federal de Minas Gerais (UFMG) instacron:UFMG |
instname_str |
Universidade Federal de Minas Gerais (UFMG) |
instacron_str |
UFMG |
institution |
UFMG |
reponame_str |
Repositório Institucional da UFMG |
collection |
Repositório Institucional da UFMG |
bitstream.url.fl_str_mv |
https://repositorio.ufmg.br/bitstream/1843/34629/2/mda.pdf https://repositorio.ufmg.br/bitstream/1843/34629/3/license.txt |
bitstream.checksum.fl_str_mv |
f8209fe4a75e797c5b4f282a18669c66 34badce4be7e31e3adb4575ae96af679 |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 |
repository.name.fl_str_mv |
Repositório Institucional da UFMG - Universidade Federal de Minas Gerais (UFMG) |
repository.mail.fl_str_mv |
|
_version_ |
1797970925704970240 |