Taming uncertainty in distributed systems with help from the network

Joshua B. Leners, Trinabh Gupta, Marcos K. Aguilera, Michael Walfish

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Network and process failures cause complexity in distributed applications. When a remote process does not respond, the application cannot tell if the process or network have failed, or if they are just slow. Without this information, applications can lose availability or correctness. To address this problem, we propose Albatross, a service that quickly reports to applications the current status of a remote process-whether it is working and reachable, or not. Albatross is targeted at data centers equipped with software defined networks (SDNs), allowing it to discover and enforce network partitions: Albatross borrows the old observation that it can be better to cause a problem than to live with uncertainty, and applies this idea to networks. When enforcing partitions, Albatross avoids disruption by disconnecting only individual processes (not entire hosts), and by allowing them to reconnect if the application chooses. We show that, under Albatross, distributed applications can bypass the complexity caused by network failures and that they become more available.

Original languageEnglish (US)
Title of host publicationProceedings of the 10th European Conference on Computer Systems, EuroSys 2015
PublisherAssociation for Computing Machinery, Inc
ISBN (Print)9781450332385
DOIs
StatePublished - Apr 17 2015
Event10th European Conference on Computer Systems, EuroSys 2015 - Bordeaux, France
Duration: Apr 21 2015Apr 24 2015

Other

Other10th European Conference on Computer Systems, EuroSys 2015
CountryFrance
CityBordeaux
Period4/21/154/24/15

Fingerprint

Uncertainty
Availability

ASJC Scopus subject areas

  • Hardware and Architecture
  • Information Systems
  • Computer Networks and Communications
  • Software
  • Control and Systems Engineering

Cite this

Leners, J. B., Gupta, T., Aguilera, M. K., & Walfish, M. (2015). Taming uncertainty in distributed systems with help from the network. In Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015 [9] Association for Computing Machinery, Inc. https://doi.org/10.1145/2741948.2741976

Taming uncertainty in distributed systems with help from the network. / Leners, Joshua B.; Gupta, Trinabh; Aguilera, Marcos K.; Walfish, Michael.

Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015. Association for Computing Machinery, Inc, 2015. 9.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Leners, JB, Gupta, T, Aguilera, MK & Walfish, M 2015, Taming uncertainty in distributed systems with help from the network. in Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015., 9, Association for Computing Machinery, Inc, 10th European Conference on Computer Systems, EuroSys 2015, Bordeaux, France, 4/21/15. https://doi.org/10.1145/2741948.2741976
Leners JB, Gupta T, Aguilera MK, Walfish M. Taming uncertainty in distributed systems with help from the network. In Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015. Association for Computing Machinery, Inc. 2015. 9 https://doi.org/10.1145/2741948.2741976
Leners, Joshua B. ; Gupta, Trinabh ; Aguilera, Marcos K. ; Walfish, Michael. / Taming uncertainty in distributed systems with help from the network. Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015. Association for Computing Machinery, Inc, 2015.
@inproceedings{3ffdd0a580264e51a6631b67dfa30b7f,
title = "Taming uncertainty in distributed systems with help from the network",
abstract = "Network and process failures cause complexity in distributed applications. When a remote process does not respond, the application cannot tell if the process or network have failed, or if they are just slow. Without this information, applications can lose availability or correctness. To address this problem, we propose Albatross, a service that quickly reports to applications the current status of a remote process-whether it is working and reachable, or not. Albatross is targeted at data centers equipped with software defined networks (SDNs), allowing it to discover and enforce network partitions: Albatross borrows the old observation that it can be better to cause a problem than to live with uncertainty, and applies this idea to networks. When enforcing partitions, Albatross avoids disruption by disconnecting only individual processes (not entire hosts), and by allowing them to reconnect if the application chooses. We show that, under Albatross, distributed applications can bypass the complexity caused by network failures and that they become more available.",
author = "Leners, {Joshua B.} and Trinabh Gupta and Aguilera, {Marcos K.} and Michael Walfish",
year = "2015",
month = "4",
day = "17",
doi = "10.1145/2741948.2741976",
language = "English (US)",
isbn = "9781450332385",
booktitle = "Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015",
publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - Taming uncertainty in distributed systems with help from the network

AU - Leners, Joshua B.

AU - Gupta, Trinabh

AU - Aguilera, Marcos K.

AU - Walfish, Michael

PY - 2015/4/17

Y1 - 2015/4/17

N2 - Network and process failures cause complexity in distributed applications. When a remote process does not respond, the application cannot tell if the process or network have failed, or if they are just slow. Without this information, applications can lose availability or correctness. To address this problem, we propose Albatross, a service that quickly reports to applications the current status of a remote process-whether it is working and reachable, or not. Albatross is targeted at data centers equipped with software defined networks (SDNs), allowing it to discover and enforce network partitions: Albatross borrows the old observation that it can be better to cause a problem than to live with uncertainty, and applies this idea to networks. When enforcing partitions, Albatross avoids disruption by disconnecting only individual processes (not entire hosts), and by allowing them to reconnect if the application chooses. We show that, under Albatross, distributed applications can bypass the complexity caused by network failures and that they become more available.

AB - Network and process failures cause complexity in distributed applications. When a remote process does not respond, the application cannot tell if the process or network have failed, or if they are just slow. Without this information, applications can lose availability or correctness. To address this problem, we propose Albatross, a service that quickly reports to applications the current status of a remote process-whether it is working and reachable, or not. Albatross is targeted at data centers equipped with software defined networks (SDNs), allowing it to discover and enforce network partitions: Albatross borrows the old observation that it can be better to cause a problem than to live with uncertainty, and applies this idea to networks. When enforcing partitions, Albatross avoids disruption by disconnecting only individual processes (not entire hosts), and by allowing them to reconnect if the application chooses. We show that, under Albatross, distributed applications can bypass the complexity caused by network failures and that they become more available.

UR - http://www.scopus.com/inward/record.url?scp=84929590674&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929590674&partnerID=8YFLogxK

U2 - 10.1145/2741948.2741976

DO - 10.1145/2741948.2741976

M3 - Conference contribution

SN - 9781450332385

BT - Proceedings of the 10th European Conference on Computer Systems, EuroSys 2015

PB - Association for Computing Machinery, Inc

ER -