Network Working Group H. Song Internet-Draft N. Zong Intended status: Standards Track Huawei Expires: April 17, 2010 Y. Yang R. Alimi Yale University October 14, 2009 DECoupled Application Data Enroute (DECADE) Problem Statement draft-song-decade-problem-statement-00 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 http://www.ietf.org/shadow.html. This Internet-Draft will expire on April 17, 2010. Copyright Notice Copyright (c) 2009 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 in effect on the date of publication of this document (http://trustee.ietf.org/license-info). Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Song, et al. Expires April 17, 2010 [Page 1] Internet-Draft DECADE Problem Statement October 2009 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. Traditional caches (e.g., P2P and Web caches) provide such storage, but they are complex (due to explicitly supporting individual P2P application protocols) and they do not allow users to manage access to content in the cache. For example, Content Providers cannot easily control access and resource usage policies to satisfy their own requirements. This document discusses the introduction of in-network storage for P2P applications, shows the need for a standard protocol for accessing this storage, and identifies the scope of this protocol. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Terminology and Concepts . . . . . . . . . . . . . . . . . . . 4 3. The Problems . . . . . . . . . . . . . . . . . . . . . . . . . 4 3.1. P2P infrastructural stress and inefficiency . . . . . . . 5 3.2. P2P cache: a complex in-network storage . . . . . . . . . 5 3.3. Ineffective integration of P2P applications and in-network storage . . . . . . . . . . . . . . . . . . . . 6 4. DECADE as an In-network Storage Capability . . . . . . . . . . 6 4.1. Data access . . . . . . . . . . . . . . . . . . . . . . . 8 4.2. Authorization . . . . . . . . . . . . . . . . . . . . . . 8 4.3. Resource control . . . . . . . . . . . . . . . . . . . . . 8 5. Usage Scenarios . . . . . . . . . . . . . . . . . . . . . . . 8 5.1. BitTorrent . . . . . . . . . . . . . . . . . . . . . . . . 8 5.2. Content Publisher . . . . . . . . . . . . . . . . . . . . 9 5.3. Data Transfer Scenarios . . . . . . . . . . . . . . . . . 10 5.3.1. Both Sender and Receiver Accounts are Used . . . . . . 10 5.3.2. Only Sender's Storage Account is Used . . . . . . . . 11 5.3.3. Only Receiver's Storage Account is Used . . . . . . . 11 5.3.4. No Storage Accounts are Used . . . . . . . . . . . . . 11 6. Security Considerations . . . . . . . . . . . . . . . . . . . 12 6.1. Denial of Service attack . . . . . . . . . . . . . . . . . 12 6.2. Copyright and Legal issues . . . . . . . . . . . . . . . . 12 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 12 8. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 12 9. References . . . . . . . . . . . . . . . . . . . . . . . . . . 13 9.1. Normative References . . . . . . . . . . . . . . . . . . . 13 9.2. Informative References . . . . . . . . . . . . . . . . . . 13 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 14 Song, et al. Expires April 17, 2010 [Page 2] Internet-Draft DECADE Problem Statement October 2009 1. Introduction P2P applications, including both P2P streaming and P2P filesharing applications, make up a large fraction of the traffic in many Internet networks today. One way to reduce bandwidth usage by P2P applications is to introduce storage capabilities in the networks. Allowing P2P applications to store and retrieve data from inside networks can reduce traffic on the last-mile uplink, as well as backbone and transit links [I-D.weaver-alto-edge-caches]. P2P caches provide in-network storage and have been deployed in some networks. But the current P2P caching architecture poses challenges to both P2P cache vendors and P2P application developers. For P2P cache vendors, it is challenging to support a number of continuously evolving P2P application protocols, due to lack of documentation, ongoing protocol changes, and rapid introduction of new features by P2P applications. For P2P applications, closed P2P caching systems limit P2P applications to effectively utilize in-network storage. In particular, P2P caches typically do not allow users to explicitly store content into in-network storage. Neither do they allow users to implement controlover the content that have have placed into the in-network storage. Both of these challenges can be effectively addressed by using an open, standard protocol to access in-network storage. P2P applications can store and retrieve content in the in-network storage, as well as control resources (e.g., bandwidth, connections) consumed by peers in a P2P application. As a simple example, a peer of a P2P application may upload to other peers through its in-network storage, saving its usage of last-mile uplink bandwidth. In this document, we distinguish between two functional components of the native P2P application protocol: signaling and data transport. Signaling includes operations such as handshaking and discovering peer and content availability. The data transport component transfers content from one peer to another. With DECADE, P2P applications can still use their native protocols for signaling and data transport. However, they may use a standard protocol for data transport incorporating in-network storage, and fall back to their native data transport protocols if in-network storage is not available or not used. In essence, an open, standard protocol to access in-network storage provides an alternative mechanism for P2P application data transport that is decoupled from P2P application control and signaling. This decoupling leads to many advantages, which is explained further in Section 4. Song, et al. Expires April 17, 2010 [Page 3] Internet-Draft DECADE Problem Statement October 2009 2. Terminology and Concepts The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119]. The following terms have special meaning in the definition of the in- network storage system. In-network Storage: A service inside a network that provides storage and bandwidth to network applications. In-network storage may reduce upload/transit/backbone traffic and improve network application performance. P2P Cache (Peer to Peer Cache): a kind of in-network storage that understands the signaling and transport of specific P2P application protocols. IAP (In-network storage Access Protocol): a standard protocol that is spoken between P2P applications and in-network storage. The protocol may also be used between entities implementing the in- network storage service. Content Publisher: An entity that originates content to consumers. 3. The Problems The emergence of peer-to-peer (P2P) as a major type of network applications (esp. P2P file sharing and streaming apps) has led to substantial opportunities. The P2P paradigm can be utilized in designing highly scalable and robust applications at low cost, compared with traditional client-server paradigms. For example, CNN [CNN] reported that P2P streaming by Octoshape played a major role in its distribution of the historical inauguration address of President Obama. PPLive, one of the largest P2P streaming vendors, is able to distribute large-scale, live streaming programs to more than 2 million users with only a handful of servers. However, P2P applications also face substantial design challenges. A particular problem facing P2P applications is the substantial stress that they place on the network infrastructure. Also, lacking of infrastructure support can lead to unstable P2P application performance during peer churns and flash crowd. Below we elaborate on the problems in more detail. Song, et al. Expires April 17, 2010 [Page 4] Internet-Draft DECADE Problem Statement October 2009 3.1. P2P infrastructural stress and inefficiency A particular problem of the P2P paradigm is the stress that P2P application traffic places on the infrastructure of Internet service providers (ISP). Multiple measurements (e.g., [ipoque]) have shown that P2P traffic has become a major type of traffic on some networks. Furthermore, network-agnostic peering leads to unnecessary traversal across network domains or spanning the backbone of a network, leading to network inefficiency [I-D.ietf-alto-problem-statement]. Recently, the IETF Application Layer Traffic Optimization (ALTO) Working Group is formed to provide P2P applications with network information so that they can perform better-than-random initial peer selection [I-D.ietf-alto-problem-statement]. However, there are limitations on what ALTO can achieve alone. For example, network information alone cannot reduce P2P traffic in access networks, as the total access upload traffic is equal to the total access download traffic in a pure P2P system. On the other hand, it is reported that P2P traffic is becoming the dominating traffic on the access networks of some networks, reaching as high as 50-60% at the down-links and 60-90% at the uplinks ([DCIA], [ICNP], [ipoque.P2P_survey.], [P2P_file_sharing]). Consequently, it becomes increasingly important to complement the ALTO effort and reduce upload access traffic, in addition to cross-domain and backbone traffic. 3.2. P2P cache: a complex in-network storage An effective technique to reduce P2P infrastructural stress and inefficiency is to introduce in-network storage. For example, in [I-D.weaver-alto-edge-caches], the author demonstrates clearly the potential benefits of introducing in-network storage to improve network efficiency and thus reduce network infrastructure stress. In the current Internet, in-network storage is introduced as P2P caches, either transparently or explicitly as a P2P peer. To provide service to a specific P2P application, the P2P cache server must support the specific signaling and transport protocols of the specific P2P application. This can lead to substantial complexity. First, there are multiple P2P applications on the Internet (e.g., BitTorrent, eMule, Flashget, and Thunder for file sharing; Abacast, Kontiki, Octoshape, PPLive, PPStream, and UUSee for P2P streaming). Consequently, a P2P cache vendor faces the challenge of supporting a large number of P2P application protocols, leading to product complexity and increased development cost. Furthermore, a specific P2P application protocol may be evolving continuously, to add new features or fix bugs. This forces a P2P Song, et al. Expires April 17, 2010 [Page 5] Internet-Draft DECADE Problem Statement October 2009 cache vendor to continuously update to track the changes of the P2P application, leading to product complexity, high cost, and low reliability. Third, many P2P applications use proprietary protocols or support end-to-end encryption. This can render P2P caches ineffective. 3.3. Ineffective integration of P2P applications and in-network storage As P2P applications evolve, it becomes increasingly clear that they will need in-network resources to provide positive user experiences. For example, multiple P2P streaming systems seek additional in- network resources during flash crowd, such as just before a major live streaming event. In asymmetric networks when the aggregated upload bandwidth of a channel cannot meet the download demand, a P2P application may seek additional in-network resources to maintain a stable system. A requirement by some P2P applications in using in-network infrastructural resources, however, is flexibility in implementing resource allocation policies. A major competitive advantage of many successful P2P systems is their substantial expertise in how to most efficiently utilize peer and infrastructural resources. For example, many live P2P systems have their specific algorithms in selecting the peers that should receive from 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 P2P applications to implement specific policies. This conforms to the end-to-end systems principle and allows innovation and satisfaction of specific business goals. Existing techniques for P2P application in-network storage lack these capabilities. 4. DECADE as an In-network Storage Capability The objective of this working group is to design DECADE, an in- network storage protocol to address the problems discussed in the preceding section. DECADE will provide access to storage and data transport services in the network to P2P applications to improve their efficiency and reduce the stress on the network infrastructure. Unlike the existing P2P caching architecture, DECADE is a standard interface for various P2P applications (both content publishers and end users) to access in-network storage. This decoupling of P2P data transport from P2P application control and signaling reduces the complexity of in- network storage services. Furthermore, DECADE provides basic access Song, et al. Expires April 17, 2010 [Page 6] Internet-Draft DECADE Problem Statement October 2009 mechanisms and allows P2P applications to implement flexible policies to create an ecosystem for application innovation and various business goals. IAP -----------+ | | | v +--------------------+ | In-network Storage | +--------------------+ ^ ^ IAP | IAP | +-------------v-+ +-v-------------+ | P2P | | Content | | application | | publishers | | clients | +---------------+ +---------------+ | ^ | | +-------------+ P2P application native protocol Figure 1 Overview of protocol interaction between DECADE elements Specifically, the main component of DECADE is the in-network storage access protocol (IAP), which is a standard, P2P-application-agnostic protocol for different P2P applications to access in-network storage. IAP defines a set of commands that P2P application elements can issue to in-network storage to store and retrieve data. IAP includes the following functionalities: (1) Data access provides read/write by users (e.g., P2P application clients and content publishers) to the corresponding in-network storage and between entities implementing the in-network storage; (2) Authorization implements access control to resources provided by in-network storage; (3) Resource control allows users to implement application policies for the corresponding in-network storage. Note that DECADE is independent of current IETF work on P2P, e.g. P2PSIP, ALTO, PPSP. For example, peers discovered by either RELOAD or ALTO can all use DECADE to share data. DECADE will focus on the problems of P2P applications, although the solution might be used in other context. Song, et al. Expires April 17, 2010 [Page 7] Internet-Draft DECADE Problem Statement October 2009 4.1. Data access P2P application clients use the protocol to read data from an in- network storage, store data to an in-network storage, or remove data from an in-network storage. The data could be of various types. 4.2. Authorization In-network storage only permits the users with authorization to access the corresponding contents. The admission could be based on user, content, time period, etc. 4.3. Resource control A user uses the protocol to manage the resources on in-network storage that can be used by other peers, e.g., the bandwidth or connections. 5. Usage Scenarios Usage scenarios are described from two perspectives. First, we introduce high-level use cases showing how P2P applications may utilize in-network storage. Second, we show how in more detail how users exchange data using IAP. 5.1. BitTorrent BitTorrent may be integrated with DECADE to be more network efficient and reduce the bandwidth consumed on ISP networks. When a BitTorrent client uploads a block to peers, the block traverses the last-mile uplink once for each peer. With DECADE, however, the BitTorrent client may upload the block to its in-network storage. Peers may retrieve the block from the in-network storage, reducing the amount of data on the last-mile uplink. We now describe in more detail how BitTorrent can utilize DECADE. For illustration, we assume that both the BitTorrent client (A) and its peer (B) utilize in-network storage. When A requests a block, it indicates that it would like the block stored in its in-network storage and provides the necessary access control. Instead of sending the 'piece' message with the desired block, peer B replies with a 'redirect' message indicating that the content should be retrieved from its own in-network storage and provides the necessary access control. If the peer B had not previously stored the content in its in-network storage, it uploads the block. A downloads the block into its own in-network storage from B's in-network storage, and finally A itself retrieves the block. Song, et al. Expires April 17, 2010 [Page 8] Internet-Draft DECADE Problem Statement October 2009 Note that this requires extensions to the BitTorrent protocol. While there are multiple ways to do so, this example assumes the native BitTorrent 'request' message is extended to carry additional information and that a new 'redirect' message is added. Upload and download to/from in-network storage uses the standard IAP protocol. This example has illustrated how utilizing DECADE can increase BitTorrent's network efficiency. First, notice that peer B does not utilize any uplink bandwidth if the block was already present in its in-network storage. Second, notice that the block is downloaded directly into A's in-network storage. When A wishes to share the block with another peer (say, peer C) that supports DECADE, it may upload directly from its in-network storage, again avoiding usage of the last-mile uplink. This technique can be applied to other P2P applications as well. Since P2P applications use a standard for communicating with in- network storage, they no longer require in-network storage to explicitly support their protocol. P2P applications retain the ability to explicitly manage which content is placed in in-network storage, as well as access and resource control polices. 5.2. Content Publisher Content Publishers may also utilize in-network storage. For example, consider a P2P live streaming application. A Content Publisher typically maintains a small number of sources, each of which distributes blocks in the current play buffer to a set of the P2P peers. Consider a case where the Content Publisher owns an in-network storage account within ISP A. If there are multiple P2P peers within ISP A, the Content Publisher may utilize DECADE to distribute content to the peers. First, the Content Publisher stores a block in the in-network storage, and then sends to each peer in ISP A the block's identifier and necessary access control. Second, each peer may then download from the Content Publisher's in-network storage. In this example, the block is distributed in a more network efficient way (the content only traverses the ISP's interdomain link once), while the Content Publisher retains explicit control over access to the content placed in its own storage. The Content Publisher can remove content from its in-network storage when it is stale or needs to be replaced, and grant access and resources to only the desired peers. Song, et al. Expires April 17, 2010 [Page 9] Internet-Draft DECADE Problem Statement October 2009 Note that Content Publishers and individual peers can each use in- network storage. For example, after downloading content from the Content Publisher's in-network storage, peers may each utilize their own in-networks storage similar to the usage scenario in Section 5.1. This can have the benefit of increased network efficiency, while Content Publishers and peers still retain control over content placed in their own in-network storage. 5.3. Data Transfer Scenarios The previous usage scenarios have utilized the ability for peers to transfer data by storing and retrieving from in-network storage. This section describes in further detail an example solution of how DECADE can provide this capability. In this section, we consider the case of a user B (the receiver) requesting data from user A (the sender). We use Sa to denote User A's storage account, and Sb to denote User B's storage account. Each user independently decides if its in-network storage account is used, so there are four cases. When a user indicates that it wishes to use its in-network storage, it provides an access control token the other user. The token is sent using the application's protocol. To simplify the illustration, we omit details of the access control from the detailed scenarios below. 5.3.1. Both Sender and Receiver Accounts are Used This scenario is illustrated in Figure 2. B first requests an object from A using the application protocol indicating it wishes the object to be stored in Sb. A responds using the application protocol indicating that B should download the object from Sa. B sends a IAP request to Sb indicating that the object should be downloaded from Sa. Sb uses IAP to download from Sa, and finally, B downloads the object from Sb (also using IAP). +-------+ IAP Get +-------+ | Sa | <-------+ Sb | +-------+ 4 +-------+ ^ \ 3 IAP Get 1 App request \ +-------+<-------------------+-------+ |User A | |User B | +-------+------------------->+-------+ 2 App response Song, et al. Expires April 17, 2010 [Page 10] Internet-Draft DECADE Problem Statement October 2009 Figure 2: Usage Scenario 1 (Sender and receiver Accounts used) 5.3.2. Only Sender's Storage Account is Used This scenario is illustrated in Figure 3. B requests an object from A using the application protocol. A responds using the application protocol indicating that B should download the object from Sa. Finally, B sends a IAP request to Sa to download the object. +-------+ | Sa | +-------+ ^ \ 3 IAP Get 1 App request \ +-------+<-------------------+-------+ |User A | |User B | +-------+------------------->+-------+ 2 App response Firgure 3: Usage Scenario 2 (Sender account used) 5.3.3. Only Receiver's Storage Account is Used This scenario is illustrated in Figure 4. B requests an object from A using the application protocol indicating that it wishes the object to be stored in Sb. A stores the object in Sb (using IAP), and responds to B (using the application protocol) that it should download from Sb. B uses IAP to download the object from Sb. +-------+ > | Sb | 3. / +-------+ IAP Store / ^ / \ 3.IAP Get / 1.App request \ +-------+<-------------------+-------+ |User A | |User B | +-------+------------------->+-------+ 2.App response Figure 4: Usage Scenario 3 (Receiver account used) 5.3.4. No Storage Accounts are Used This scenario is illustrated in Figure 5. In this scenario, the application protocol is used directly to send data. This scenario applies with one of the peers does not support IAP, or neither of the peers are using in-network storage. Song, et al. Expires April 17, 2010 [Page 11] Internet-Draft DECADE Problem Statement October 2009 1.App request +-------+<-------------------+-------+ |User A | ... ... ... |User B | +-------+------------------->+-------+ 2.data transfer Figure 5: Usage Scenario 4 ( No storage accounts used) 6. Security Considerations There are multiple security considerations. We focus on two in this section. 6.1. Denial of Service attack Without access control or resource control, an attacker can try to consume a large portion of the in-network storage, or exhaust the connections of the in-network storage to commit a Denial of Service (DoS) attack. Thus, access control and resource control mechanisms are mandatory. A resource control mechanism should be used to allow a user to allocate the resource in its in-network storage account to be utilized by other clients. 6.2. Copyright and Legal issues Copyright and other laws may prevent the distribution of certain content in various localities. While in-network storage operators may adopt system-wide ingress or egress filters to implement necessary policies for storing or retrieving content, the specification and implementation of such policies (e.g., filtering and DRM) is outside of the scope of this working group. 7. IANA Considerations There is no IANA consideration with this document. 8. Acknowledgments We would like to thank the following people for contributing to the discussion of this document: Kar Ann Chew Richard Woundy Song, et al. Expires April 17, 2010 [Page 12] Internet-Draft DECADE Problem Statement October 2009 Yunfei Zhang Yu-shun Wang David Bryan Roni Even Yingjie Gu Hongqiang Law. 9. References 9.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. 9.2. Informative References [ipoque.com] "http://www.ipoque.com/resources/internet-studies/ internet-study-2008_2009". [I-D.ietf-alto-problem-statement] Seedorf, J. and E. Burger, "Application-Layer Traffic Optimization (ALTO) Problem Statement", draft-ietf-alto-problem-statement-04 (work in progress), September 2009. [I-D.weaver-alto-edge-caches] Weaver, N., "Peer to Peer Localization Services and Edge Caches", draft-weaver-alto-edge-caches-00 (work in progress), March 2009. [DCIA] http://www.dcia.info, "Distributed Computing Industry Association". [ipoque.P2P_survey.] "Emerging Technologies Conference at MIT", Sept. 2007. [P2P_file_sharing] Parker, A., "The true picture of peer-to-peer filesharing", July 2004. [ICNP] Wu, H., "Challenges and opportunities of Internet Song, et al. Expires April 17, 2010 [Page 13] Internet-Draft DECADE Problem Statement October 2009 developments in China, ICNP 2007 Keynote", Oct. 2007. Authors' Addresses Song Haibin Huawei Baixia Road No. 91 Nanjing, Jiangsu Province 210001 P.R.China Phone: +86-25-84565867 Email: melodysong@huawei.com Ning Zong Huawei Baixia Road No. 91 Nanjing, Jiangsu Province 210001 P.R.China Phone: +86-25-84565866 Email: zongning@huawei.com Y. Richard Yang Yale University Email: yry@cs.yale.edu Richard Alimi Yale University Email: richard.alimi@yale.edu Song, et al. Expires April 17, 2010 [Page 14]