DECADE R. Alimi
Internet-Draft Google
Intended status: Informational Y. Yang
Expires: April 21, 2011 Yale University
A. Rahman
InterDigital Communications, LLC
D. Kutscher
NEC Laboratories Europe
L. Chen
H. Liu
Yale University
October 18, 2010
DECADE Architecture
draft-alimi-decade-arch-00
Abstract
Peer-to-peer (P2P) applications have become widely used on the
Internet today and make up a large portion of the traffic in many
networks. In P2P applications, one technique for reducing the total
amount of P2P traffic is to introduce storage capabilities within the
network. The DECADE Working Group has been formed with the goal of
developing an architecture to provide this capability. This
documents presents an architecture, discusses the underlying
principles and identifies core components and protocols supporting
the architecture.
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at
Alimi, et al. Expires April 21, 2011 [Page 1]
Internet-Draft DECADE Architecture October 2010
http://www.ietf.org/shadow.html.
This Internet-Draft will expire on April 21, 2011.
Copyright Notice
Copyright (c) 2010 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the BSD License.
Alimi, et al. Expires April 21, 2011 [Page 2]
Internet-Draft DECADE Architecture October 2010
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4
2. Entities . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1. DECADE Storage Servers . . . . . . . . . . . . . . . . . . 5
2.2. DECADE Storage Provider . . . . . . . . . . . . . . . . . 5
2.3. Content Distribution Application . . . . . . . . . . . . . 5
2.4. DECADE Content Providers . . . . . . . . . . . . . . . . . 5
2.5. DECADE Content Consumers . . . . . . . . . . . . . . . . . 5
2.6. End-Point . . . . . . . . . . . . . . . . . . . . . . . . 6
3. Architectural Principles . . . . . . . . . . . . . . . . . . . 6
3.1. Decoupled Control and Data Planes . . . . . . . . . . . . 6
3.2. Immutable Data Objects . . . . . . . . . . . . . . . . . . 7
3.3. Accessing Data Objects . . . . . . . . . . . . . . . . . . 8
3.4. Data Object Identifiers . . . . . . . . . . . . . . . . . 8
3.5. Explicit Control . . . . . . . . . . . . . . . . . . . . . 8
3.6. Resource and Data Access Control through User
Delegation . . . . . . . . . . . . . . . . . . . . . . . . 8
3.6.1. Resource Allocation . . . . . . . . . . . . . . . . . 9
3.6.2. User Delegations . . . . . . . . . . . . . . . . . . . 9
4. System Components . . . . . . . . . . . . . . . . . . . . . . 9
5. Reference Architecture . . . . . . . . . . . . . . . . . . . . 10
6. Security Considerations . . . . . . . . . . . . . . . . . . . 10
7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10
8. Informative References . . . . . . . . . . . . . . . . . . . . 10
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 10
Alimi, et al. Expires April 21, 2011 [Page 3]
Internet-Draft DECADE Architecture October 2010
1. Introduction
Peer-to-peer (P2P) applications have become widely used on the
Internet today to distribute contents, and they contribute a large
portion of the traffic in many networks. The DECADE Working Group
has been formed with the goal of developing an architecture to
introduce in-network storage to be used by such applications, to
achieve more efficient content distribution. Specifically, in many
subscriber networks, it is typically more expensive to upgrade
network equipment in the "last-mile", because it can involve
replacing equipment and upgrading wiring at individual homes,
businesses, and devices such as DSLAMs and CMTSs. Thus, it can be
cheaper to upgrade core infrastructure involving fewer components
that are shared by many subscribers. See
[I-D.ietf-decade-problem-statement] for a more complete discussion of
the problem domain and general discussion of the capabilities to be
provided by DECADE.
This document presents a potential architecture of providing in-
network storage that can be integrated into content distribution
applications. The primary focus is P2P-based content distribution,
but the architecture may be useful to other applications with similar
characteristics and requirements. In particular, content
distribution applications that may split data into smaller pieces for
distribution may be able to utilize DECADE.
The design philosophy of the DECADE architecture is to provide only
the core functionality that is needed for applications to make use of
in-network storage. With such core functionality, the protocol may
be simple and easier to support by storage providers. If more
complex functionality is needed by a certain application or class of
applications, it may be layered on top of the DECADE protocol.
The DECADE protocol will leverage existing transport and application
layer protocols and will be designed to work with a small set of
alternative IETF protocols.
This document proceeds in two steps. First, it details the core
architectural principles that can guide the DECADE design. Next,
given these core principles, this document presents the core
components of the DECADE architecture and identifies usage of
existing protocols and where there is a need for new protocol
development.
This document will be updated to track the progress of the DECADE
survey [I-D.ietf-decade-survey] and requirements [I-D.gu-decade-reqs]
drafts.
Alimi, et al. Expires April 21, 2011 [Page 4]
Internet-Draft DECADE Architecture October 2010
2. Entities
2.1. DECADE Storage Servers
DECADE storage servers are operated by DECADE storage providers and
provide the DECADE functionality as specified in this memo, including
mechanisms to store, retrieve and manage data. A storage provider
will typically operate many storage servers.
2.2. DECADE Storage Provider
A DECADE in-storage provider deploys and/or manages DECADE servers
within a network. Storage providers may also own or manage the
network in which the DECADE servers are deployed.
A storage provider, possibly in cooperation with one or more network
providers, determines deployment locations for DECADE servers and
determines the available resources for each.
2.3. Content Distribution Application
A content distribution application is developed by an Application
Developer and installed on end-hosts. End-hosts may be machines
managed by the Application Developer itself, a Content Provider, or
an End-User.
2.4. DECADE Content Providers
DECADE content providers access DECADE storage servers (by way of a
DECADE client) to upload and manage data in the context of an context
distribution application. A content provider can access one or more
storage servers. A content provider may be a single process or a
distributed application (in a P2P scenario).
2.5. DECADE Content Consumers
DECADE content consumers access storage servers (by way of a DECADE
client) to download data that has previously been stored by a content
provider in the context of a content distribution application. A
content consumer can access one more storage servers. A content
consumer may be a single process or a distributed application (in a
P2P scenario). An instance of a distributed application, such as a
P2P application, may both provide content to and consume content from
DECADE storage servers.
Alimi, et al. Expires April 21, 2011 [Page 5]
Internet-Draft DECADE Architecture October 2010
2.6. End-Point
An End-Point is an instance of a Content Distribution Application
that includes a DECADE client. A particular End-Point may be a
DECADE Content Provider, DECADE Content Consumer, or both.
An End-Point need not be an active member of a "swarm" to interact
with the DECADE storage system. That is, an End-Point may interact
with the DECADE storage servers as an offline activity.
3. Architectural Principles
We identify the following key principles.
3.1. Decoupled Control and Data Planes
The DECADE infrastructure is intended to support multiple content
distribution applications. A complete content distribution
application implements of a set of control functions including
content search, indexing and collection, access control, ad
insertion, replication, request routing, and QoS scheduling.
Different content distribution applications can have unique
considerations designing the control and signaling functions. For
example, a major competitive advantage of many successful P2P systems
is their substantial expertise in how to most efficiently utilize
peer and infrastructural resources. Many live P2P systems have their
specific algorithms in selecting the peers that behave as the more
stable, higher-bandwidth sources. They continue to fine-tune such
algorithms. In other words, in-network storage should export basic
mechanisms and allow as much flexibility as possible to the control
planes to implement specific policies. This conforms to the end-to-
end systems principle and allows innovation and satisfaction of
specific business goals.
Specifically, in the DECADE architecture, the control plane focuses
on the application-specific, complex, and/or processing intensive
functions while the data plane provides storage and data transport
functions.
o Control plane: Signals on which data are downloaded to whom, from
where, at what time, and with what quality-of-service (e.g.,
bandwidth). It also provides higher layer meta-data management
functions such as defining the sequence of data blocks forming a
higher layer content object. These are behaviors designed and
implemented by the Application.
Alimi, et al. Expires April 21, 2011 [Page 6]
Internet-Draft DECADE Architecture October 2010
o Data plane: Stores and transfers data as instructed by the
Application's Control Plane.
Decoupling control plane and data plane is not new. For example,
OpenFlow is an implementation of this principle for Internet routing,
where the computation of the forwarding table and the application of
the forwarding table are separated. Google File System applies the
principle to file system design, by utilizing the Master to handle
the meta-data management, and the chunk servers to handle the data
plane (i.e., read and write of chunks of data). NFS4 also implements
this principle.
Note that applications may have different Data Plane implementations
in order to support particular requirements (e.g., low latency). In
order to provide interoperability, the DECADE architecture does not
intend to enable arbitrary data transport protocols between DECADE
servers. However, the architecture should allow for multiple data
transport protocols to be used.
Also note that although an application's existing control plane
functions remain implemented within the application, the particular
implementation may need to be adjusted to support DECADE.
3.2. Immutable Data Objects
A property of bulk contents to be distributed is that they typically
are immutable -- once a piece of content is generated, it is
typically not modified. It is not common that bulk contents such as
video frames and images need to be modified after distribution.
Many content distribution applications divide content objects into
blocks for two reasons: (1) different blocks may be fetched from
different content sources in parallel, and (2) individual blocks may
be recovered from an alternate content source. Typically,
applications use a block size larger than a single packet in order to
reduce control overhead.
Common applications whose content matches this model include P2P
streaming (live and video-on-demand) and P2P file-sharing content.
However, other types of applications may additionally match this
model.
DECADE adopts a design in which immutable data objects may be stored
at a storage server. Applications may consider existing blocks as
DECADE data objects, or or they may adjust block sizes before storing
in a DECADE server.
Focusing on immutable data blocks in the data plane can substantially
Alimi, et al. Expires April 21, 2011 [Page 7]
Internet-Draft DECADE Architecture October 2010
simplify the data plane design, since consistency requirements can be
relaxed. It also allows effective reuse of data blocks and de-
duplication of redundant data.
Note that immutable content may still be deleted. If applications
require particular content to be modified, one possibility is to
delete (or revoke access to) the existing content and then distribute
a different version.
3.3. Accessing Data Objects
The DECADE data access protocol allows clients to store and retrieve
data fragments ("data objects") of arbitrary size. The DECADE
architecture is agnostic to application-specific framing conventions,
unfiform fragment lengths etc.
3.4. Data Object Identifiers
Objects that are stored in a DECADE storage server can be accessed by
DECADE content consumers by a resource identifiers that has been
assigned within a certain application context.
Because a DECADE content consumer can access more than one storage
server within a single application context, a data object that is
replicated across different storage servers managed by a DECADE
storage provider, can be accessed by a single identifier.
3.5. Explicit Control
To support the functions of an application's control plane,
applications must be able to know and control which data is stored at
particular locations. Thus, in contrast with content caches,
applications are given explicit control over the placement (selection
of a a DECADE server), deletion (or expiration policy), and access
control for stored data.
Consider deletion/expiration policy as a simple example.
Applications may require a DECADE server to store content for a
relatively short period of time (e.g. for live-streaming data) or may
need to store content long term (e.g., for video-on-demand).
3.6. Resource and Data Access Control through User Delegation
DECADE provides a shared infrastructure to be used by multiple
tenants of multiple content distribution applications. Thus, it
needs to provide both resource and data access control.
Alimi, et al. Expires April 21, 2011 [Page 8]
Internet-Draft DECADE Architecture October 2010
3.6.1. Resource Allocation
There are two primary interacting entities in the DECADE
architecture. First, Storage Providers control where DECADE storage
servers are provisioned and their total available resources. Second,
Applications executing on end-points control data transfers amongst
available DECADE servers and between DECADE servers and end-points.
A form of isolation is required to enable concurrently-running
Applications to each explicitly manage their own content and share of
resources at the available servers.
Management of the resources at a server are delegated by a Storage
Provider to one or more applications. Applications are able to
explicitly and independently manage their own share of resources.
3.6.2. User Delegations
Storage providers have the ability to explicitly manage the entities
allowed to utilize the resources at a DECADE server. This capability
is needed for reasons such as capacity-planning and legal
considerations in certain deployment scenarios.
To provide a scalable way to manage applications granted resources at
a DECADE server, a layer of indirection is added. Instead of
granting resources to an application, the DECADE server grants a
share of the resources to a user. The user may in turn share the
granted resources amongst multiple applications. The share of
resources granted by a storage provider is called a User Delegation.
A User Delegation may be granted to an end-user (e.g., an ISP
subscriber), a Content Provider, or an Application Provider. A
particular instance of an application may make use of the storage
resources:
o granted to the end-user (with the end-user's permission),
o granted to the Content Provider (with the Content Provider's
permission>, and/or
o granted to the Application Provider.
4. System Components
The current version of the document has primarily focused on the
architectural principles. The detailed system components will be
discussed in the next document revision.
Alimi, et al. Expires April 21, 2011 [Page 9]
Internet-Draft DECADE Architecture October 2010
5. Reference Architecture
The current version of the document has primarily focused on the
architectural principles. The detailed reference architecture will
be discussed in the next document revision.
6. Security Considerations
This document currently does not contain any security considerations
beyond those mentioned in [I-D.ietf-decade-problem-statement].
7. IANA Considerations
This document does not have any IANA considerations.
8. Informative References
[I-D.ietf-decade-problem-statement]
Yongchao, S., Zong, N., Yang, Y., and R. Alimi, "DECoupled
Application Data Enroute (DECADE) Problem Statement",
draft-ietf-decade-problem-statement-00 (work in progress),
August 2010.
[I-D.ietf-decade-survey]
Alimi, R., Rahman, A., and Y. Yang, "A Survey of In-
network Storage Systems", draft-ietf-decade-survey-00
(work in progress), September 2010.
[I-D.gu-decade-reqs]
Yingjie, G., Bryan, D., Yang, Y., and R. Alimi, "DECADE
Requirements", draft-gu-decade-reqs-05 (work in progress),
July 2010.
Authors' Addresses
Richard Alimi
Google
Email: ralimi@google.com
Alimi, et al. Expires April 21, 2011 [Page 10]
Internet-Draft DECADE Architecture October 2010
Y. Richard Yang
Yale University
Email: yry@cs.yale.edu
Akbar Rahman
InterDigital Communications, LLC
Email: akbar.rahman@interdigital.com
Dirk Kutscher
NEC Laboratories Europe
Email: dirk.kutscher@neclab.eu
Lijiang Chen
Yale University
Email: lijiang.chen@yale.edu
Hongqiang Liu
Yale University
Email: hongqiang.liu@yale.edu
Alimi, et al. Expires April 21, 2011 [Page 11]