Changes between Version 46 and Version 47 of coin


Ignore:
Timestamp:
Feb 19, 2019, 5:16:32 AM (9 months ago)
Author:
marie@…
Comment:

Removed a lot of the background and focused on the new charter in preparation for Prague.

Legend:

Unmodified
Added
Removed
Modified
  • coin

    v46 v47  
    1 = Computing in the Network (COIN) =
     1= Proposed IRTF Research Group: Computing in the Network (COIN) =
     2
     3'''Co-chairs'''
     4
     5Jianfei (Jeffrey) He (Huawei) jeffrey.he@huawei.com
     6Marie-Jose Montpetit (TriangleVideo) marie@mjmontpetit.com
     7Eve Schooler (Intel) eve.m.schooler@intel.com
    28
    39'''Mailing list'''
     
    713Subscribe at: https://www.ietf.org/mailman/listinfo/Coin
    814
     15= Charter =
    916
    10 '''Introduction (to be updated)'''
     17The COIN research group (COINRG) provides a forum for the presentation, discussion, and analysis of research addressing to what extent, and how, computing and networking technologies can be merged. The scope includes architectures, protocols, and implementations that encompass both of these technology areas. The primary goals are to improve network and application performance, agility, security, and privacy by better integration. COIN will foster research into solutions that measurably improve the internet by blending how routing/forwarding, packet processing, computing, and storage are realized. Work involving both classic “control plane” and “data plane” technologies are in scope.
    1118
    12 The integration computing and networking seems natural and has been widely investigated and applied at several network layers in the past. Most notably, “Active Networking” research in the 1990s explored approaches for allowing packets and datagrams flowing through a network to modify the behavior of the network itself. This could be done at several layers, e.g., enabling/modifying transport protocol behavior, configuring or programming link layer functionality upon connection establishment etc.
     19COIN will address a wide variety of system scales, ranging from large data center networks (DCN) to small edge computing and networking complexes. Of particular interest is work that explores this continuum and addresses how cloud computing languages and tools might evolve (or not) to encompass the full range of scale, deployment, and physical realization.
    1320
    14 We are experiencing a new wave of convergence between networking and computing, triggered not only by the softwarization of networking functions (SDN, NFV) but also by the evolution of the network architecture itself. The move to the edge and distributed computing/networking is also encouraging the development of local networking and computing facilities to support low delay and low loss services that are emerging from AR/VR, autonomous vehicles, V2X and intelligent/smart cities.
     21A major focus will also be on bridging the current divide between the way server computing complexes are programmed and how switches and other smart networking devices such as NICs are programmed. While current state-of-the art in programmability via P4 and other languages/systems has resulted in improved functionality in switches, richer programmability is likely needed to support important workloads, such as edge network analytics, machine learning, and data fusion for cyber-physical systems. Such applications not only need access to more general-purpose languages and underlying operating system facilities, but also need to accommodate local and remote caches, dynamic control points, and various forms of data stewardship. The multiple “touch points” and attendant data transformations raise important research issues in security, privacy and data provenance, as a simple “end-to-end” two party model of communication does not capture the richness of the systems people envision. Research concerning how the existing layering of protocols is affected by these considerations is also in scope. COIN will foster discussion on what should (or should not) be exposed across layers and APIs, including parameters that might affect QoS/QoE, orchestration dynamics, and mobility.
    1522
    16 Recent research in network data plane programmability has also enabled new ways for relaxing the boundaries between strictly network layer and application layer programmability. For example, switch programming abstractions such as P4 and more powerful programmable data center switch platforms enable the implementation of different support functions for application layers entities, supporting applications such as DNN (Deep Neural Network) training, frontend KV (Key-Value) caching for skewed and dynamic workloads and high-performance consensus protocols such as
    17 Paxos.
    18 
    19 In addition, there are scalable stream processing frameworks such as Apache Spark or Apache Flink that apply programmed functions on data flowing in a distributed system. These platforms are typically concerned with guaranteeing certain semantics and providing high reliability and performance by orchestrating the set of functions accordingly. Such distributed processing platforms are overlays (from a network layer perspective) but also have to deal with flow/congestion control etc. on their respective layer.
    20 
    21 In parallel, there are approaches for connecting so-called network functions (such as NFV -- Network Functions Virtualization) and derived/related computing/networking models such as CDN and Edge Computing. They are mostly concerned with setting up and maintaining overlays and virtual networks between application logic in virtual machines but also corporate element of local processing and computing.
    22 
    23 Consequently the idea of a “programmable” network is central to the evolution of the Internet and for the support of emerging applications and services.
     23The ability to deploy, manage and evolve a system with both data center and edge components brings new challenges to how complex applications are orchestrated, both individually and in mutual competition for resources. The RG hopes to foster work on orchestration of multi-datacenter and edge resources through both proposals to extend existing approaches (e.g., Kubernetes, Swarm) as well as via novel approaches. Special attention will be paid to use-case-driven requirements gathered from next-generation applications and services (e.g., video streaming, immersive AR/VR, autonomous/connected vehicles, industrial IoT). In doing so, COIN expects to closely follow and coordinate with the evolving research in distributed computing and new network paradigms, for instance, information-centric networking,
    2424
    2525
    26 '''Motivation'''
     26'''Scope'''
    2727
    28 Network programmability provides new opportunities to enhance performance and availability of network as well as to develop new types of networked applications and systems. Looking at different research project that address computing and networking, the following questions emerge:
     28(1) Improvements to the programmability of network devices in order to bring more of the richness of distributed computing technologies to in-network devices.
    2929
    30 1) Are there common principles, abstractions and mechanisms that can be applied across the range of computing/networking elements?
     30(2) Use case driven requirements analysis: examine cloud computing as a continuum from the data center to edge networks and beyond incorporating the use of programmable switches and NICs. Identify the potential benefits to these networks by employing in-network functionality, including but not limited to compute, cache, manage, control, etc.
    3131
    32 2) What are research avenues and relevant considerations for COIN, in particular with respect to previous projects such as active networking and the agreed-on end-to-end-arguments?
     32(3) Novel architectures, data-plane abstractions, and new network protocol designs to efficiently federate decentralized computing resources across the infrastructure regardless of where in the network the compute is placed (the DC, the core, the edge, and even in the end-user devices).
    3333
    34 3) Many of the computing/networking systems developed for and by the networking community have assumed the existence of running applications that need to be connected
    35 directly, run on overlays or using service chains. Is there potential to include design patterns from the distributed computing and applications community that would allow considering both networking and computing for optimizing the layout of processing functions and distribution of data?
     34(4)  Privacy and security mechanisms required or enabled by in-network compute and their effect on various protocol layers.
    3635
    37 '''Research Challenges'''
    38 
    39 The research questions that the COIN group wants to address include but are not limited to:
    40 
    41 (1) Even within the traditional and "end-to-end argument", will distributed computing in the network provide enough motivation and benefits to justify the introduction of non-forwarding functions into the network?
    42 
    43 
    44 (2) Will forwarding function be eventually be integrated into the computing paradigm for example ML for route determination?
    45 
    46 
    47 (3) To which level must the abstraction of the programmable data plane be for a network with non-forwarding functions? With new functionalities, the network an infrastructure, need to be decoupled from some applications so that it can be kept stable with permanently evolving applications.
    48 
    49 
    50 (4) What will the impact of these in-network functions on end-to-end transport protocols and security? Will transport start being hop-by-hop?
    51 
    52 
    53 (5) With the network as a database what will be the impact on the privacy of users' data and identities.
    54 
    55 
    56 (6) What are the economic, social and environmental incentives for the network to add new computing capabilities and resources in an open ecosystem.
    57 
    58 
    59 
    60 '''COIN Objectives'''
    61 
    62 The main objectives of the COIN RG include:
    63 
    64 1)      Understanding the use cases and different types of network programmability and their different characteristics (for example, DC switch programmability vs. distributed/edge computing).
    65 
    66 
    67 2)      Investigating architectural questions such as system architecture and protocol designs for in-network computing, for example interactions of data and control planes as well as overall system and protocol security.
    68 
    69 
    70 3)      Understanding relationship to and impact on existing Internet protocols (transport, traffic steering) and frameworks (security, management).
    71 
    72 
    73 4)      Developing common terminology, concepts and potentially system elements such as data plane protocols and management concepts.
    74 
    75 
    76 5)      Providing guidance for potential future IETF work on distributed and in-network computing.
    77 
    78 
    79 = Draft Charter TO BE DISCUSSED =
    80 
    81 The COIN research group wants to explore the research on how to make use of new programmable data planes and distributed computing to introduce non-forwarding functions into networks and functional federation to improve network and application performance and user experience.
    82 
    83 In order to achieve this goal the methodology will include specific future-looking use cases with their outcomes, the trade-offs between the benefits from the new functionality in network and the extra cost to network devices, related research on edge computing applications that benefit from programmability and research on applications that could be moved into the network to provide added functionality. Use cases will include the collaboration between centralized and controlled environments like DCN, and the widely distributed networks characterized by edge/pervasive computing. While it is not mandatory it is hoped that later in the RG's work the combination of both approaches in a common architecture may lead to common protocols. With it's focus on research in the overall field of computing in the network the RG will not address a specific architecture, system or protocol but will assume that for example specific solutions in COIN may have the equivalent of a transport layer to consider both network and compute resources jointly.
    84 
    85 The use cases and related research may lead to new architecture and layering design with comparison to traditional architectures in terms of complexity, performance and cost and create incentivesthe research into new abstraction of the data plane and the development of some potential new protocols. Finally the impacts of COIN on transport protocols, security and privacy in different environments, and the incentives for both the network providers to provide the capabilities and the application developers to use them will also be investigated.
    86 
    87 ''Scope''
    88 
    89 (1)     Use case analysis/targeted research: DCN, edge networks, IoT networks etc. and the potential benefits to these networks from in-network non-forwarding functions like compute, cache, manage etc.
    90 
    91 (2)     Research on solutions to use current and coming programmable network devices to implement non-forwarding functions and demonstrate the relationship between performance and benefit/gains.
    92 
    93 (3)     Research on novel architectures, new data-plane abstraction and new protocol designs that make full use of the constrained compute and cache capabilities in programmable network devices and how to expand them,
    94 
    95 (4)     Research on novel architectures, new data-plane abstraction and new network protocol designs that can help to efficiently use the decentralized computing resources, inside the network devices in the DC, the core and the edge, or even in the end-user devices.
    96 
    97 (5)     Research on new transport protocol and new privacy and security mechanisms enabled by in-network non-forwarding functions.
    98 
    99 (6)     Research on incentive mechanisms to encourage both the network providers to provide the capabilities and the application developers to use them.
    100 
    101 Hence, COIN wants to focus on the challenges facing computing in the network across all network segments and identify how these challenges are met in current and future research. Moreover, the specific use cases will link the challenges to some future network architectures and implementations.
    102 
    103 ''Outcomes''
     36''Outcomes - to be discussed''
    10437
    10538COIN  wants to build a forum to explore and discuss how the network architectures and protocols will adapt to the introduction of distributed systems and decentralized computing resources. Hence the following outcomes are proposed:
     
    12255In Network Computing Enablers for Extended Reality - draft-montpetit-coin-XR-01 available at https://datatracker.ietf.org/doc/draft-montpetit-coin-xr/  (errata on v00).
    12356
     57
     58== Next meeting ==
     59
     60During IEFF 104 March 24-29: specific date and time to be specified.
    12461
    12562== References ==