2008
DOI: 10.1145/1402946.1402966
|View full text |Cite
|
Sign up to set email alerts
|

A policy-aware switching layer for data centers

Abstract: Today's data centers deploy a variety of middleboxes (e.g., firewalls, load balancers and SSL offloaders) to protect, manage and improve the performance of the applications and services they run. Unfortunately, existing networks provide limited support for middleboxes. Administrators typically overload layer-2 path selection mechanisms to make sure that traffic traverses the desired sequence of middleboxes. These ad-hoc practices result in a data center network that is hard to configure, upgrade and maintain, … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
59
0

Year Published

2011
2011
2019
2019

Publication Types

Select...
6
3

Relationship

0
9

Authors

Journals

citations
Cited by 108 publications
(59 citation statements)
references
References 14 publications
0
59
0
Order By: Relevance
“…Joseph et al [5] propose to use a layer-2 switching layer for policy enforcement. The recent works of Steering [6] and Simple [7] employ the centralized control paradigm in SDN to route traffic to different services.…”
Section: Related Workmentioning
confidence: 99%
See 1 more Smart Citation
“…Joseph et al [5] propose to use a layer-2 switching layer for policy enforcement. The recent works of Steering [6] and Simple [7] employ the centralized control paradigm in SDN to route traffic to different services.…”
Section: Related Workmentioning
confidence: 99%
“…We propose to introduce an optical steering domain in addition to the conventional packet domain to steer large aggregate flows through core NFs [4]. Our work is complementary with the packet-based works [5]- [8].…”
Section: Related Workmentioning
confidence: 99%
“…The first allows the middleboxes to be placed on the shortest path, however it requires infrastructural changes and result in poor flexibility as the policy will be applied to all traffic. The second allows more flexibility as an arbitrary policy chain can be configured through custom routing entries, at the cost of a longer path and overloaded routing tables, making it hard to configure and maintain [16]. In our approach, we reroute traffic to the GLANF Server hosting the relevant network function.…”
Section: A Traffic Routingmentioning
confidence: 99%
“…The most relevant work is the policy-aware switch (pswitch) proposed in [12]. In this design, a number of middleboxes are plugged into a pswitch that is deployed in-path.…”
Section: Related Workmentioning
confidence: 99%