INTERNET-DRAFT A. Palanivelan Intended Status: Experimental Verizon Labs Expires: Apr 19, 2016 Chetan Harsha Verizon Labs Senthil Sivakumar Cisco Systems Oct 17, 2015 MPTCP Enhancement Opportunities draft-palanivelanchetansenthil-mptcp-enhancements-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/1id-abstracts.html The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html Copyright and License Notice Copyright (c) 2015 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 Simplified BSD License. Palanivelan,Chetan,. Expires Apr 19,2016 [Page 1] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 Abstract MPTCP Intends to address a wide range of issues, with minimal implementation tweaks. Though this works in a range of use cases,there are some use cases, where some standard implementation recommendations could help. The Purpose of this draft is to document Opportunities, where Enhancements to MPTCP can translate to more wider deployments. Table of Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . 3 2 MPTCP Enhancement Opportunities - End user Use cases . . . . . . 3 2.1 Short Flows vs Long Flows . . . . . . . . . . . . . . . . . 3 2.2 Application based Path selection and Adaptive buffering . . 4 2.3 Path Selection Enhancements . . . . . . . . . . . . . . . . 4 2.4 Optimal number of paths . . . . . . . . . . . . . . . . . . 5 3 UseCase Scenarios (Simulated in Lab) and Results . . . . . . . . 6 3.1 MPTCP Enabled Client Uploads data from Non-MPTCP Capable Server . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.2 MPTCP Enabled Client Uploads data from MPTCP Enabled Server . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.3 MPTCP Enabled Client Uploads data from Non-MPTCP Capable Server with Intermediate MPTCP Enabled devices (proxy?) . . 6 3.4 MPTCP Enabled Client Uploads data from MPTCP Enabled Server with Intermediate MPTCP Enabled devices (proxy?) . . 6 3.5 MPTCP Enabled Client Downloads data from Non-MPTCP Capable Server . . . . . . . . . . . . . . . . . . . . . . 7 3.6 MPTCP Enabled Client Downloads data from MPTCP Enabled Server . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3.7 MPTCP Enabled Client Downloads data from Non-MPTCP Capable Server with Intermediate MPTCP Enabled devices (proxy?) . 7 3.8 MPTCP Enabled Client Downloads data from MPTCP Enabled Server with Intermediate MPTCP Enabled devices (proxy?) . . 7 4 Security Considerations . . . . . . . . . . . . . . . . . . . . 8 5 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 8 6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 6.1 Normative References . . . . . . . . . . . . . . . . . . . 8 6.2 Informative References . . . . . . . . . . . . . . . . . . 8 Author's Addresses . . . . . . . . . . . . . . . . . . . . . . . . 8 Palanivelan,Chetan,. Expires Apr 19,2016 [Page 2] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 1 Introduction The Scope of the use cases discussed is limited to impact on end-user experience only and recommended updates at SP (PE Router). The initial versions of this draft would document findings from tests covering various end-user use cases in detail, that presents mptcp enhancement opportunities. The later versions of the document would strive to provide solutions for the documented usecase scenarios. 1.1 Terminology 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 RFC 2119 [RFC2119]. 2 MPTCP Enhancement Opportunities - End user Use cases 2.1 Short Flows vs Long Flows Internet traffic MUST have Security, Throughput, Reliability,.. taken care across different network conditions, modes of access and flows. Data access can be categorized into short or long flows. Too many Small Flows => Higher Number of Transactions. But, much less Bandwidth Consumption Can we achieve Low latency for short flows? Average completion of flow with mptcp can be higher than completion time without mptcp With Bunch of Short Flows, MPTCP may negatively impact throughput Even a single lost packet can force an entire connection to wait for an RTO. Far Lesser Long Flows => Lesser Number of Transactions. But, higher Bandwidth Consumption Can we achieve higher Throughput for Long Flows Without compromising on performance? How do we maintain Reliability? How do we manage tolerance to sudden and high bursts of traffic? In Summary, Both long and short flows are important from the enduser perspective. We need to come up with appropriate definition and clear demarcation for short and long flows, from MPTCP Perspective. These need be dealt differently (Probably with multiple profiles). Palanivelan,Chetan,. Expires Apr 19,2016 [Page 3] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 2.2 Application based Path selection and Adaptive buffering How much of benefit it would be when we consider different type of applications, for better mptcp profiling. Typical internet applications are categorized as Elastic and InElastic. Elastic vs Inelastic Applications..How does it matter to MPTCP? MPTCP performance is impacted : When the size of the receive buffer is limited. Path with high RTT may result in the receive buffer size growing beyond the allowed maximum Diversified RTT Different ways of handling packets => Better Performance. In Summary, Application based Path Selection and Adaptive Buffering can help with the above scenarios. Tweaking the buffer sizes based on the type of application and/or network condition can positively impact the flows. 2.3 Path Selection Enhancements Path Selection is one of the important part of MPTCP. Though there are existing tools that help diagnose issues in the path, there still is scope to fine tune it further flexible based on certain factors. Usecases where MPTCP path selection can be enhanced: For High packet loss and High latency networks? Multiple profiles to dynamically switch (move across) the networks? Roaming scenarios In Summary, The best optimal path is ever changing in the Internet. Frequent switching may cause unnecessary overheads and can impact performance. Enhanced yet controlled Path Selection and Path Switching can help get better performance out of the network. Palanivelan,Chetan,. Expires Apr 19,2016 [Page 4] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 2.4 Optimal number of paths The best and effective path selection is critical to the effect of MPTCP for the client application. How about the optimal number of sub-flows? Can we improve client experience by controlling number of sub flows based on certain factors? Controlling the number of sub flows getting created: How many is too many? Can this be controlled? What Inputs to Consider? Based on Network Characteristics Historic data (region wise) In Summary, MPTCP being not too strict as well as not too flexible, Certain profiling based on detailed analysis of data can positively impact MPTCP experience Palanivelan,Chetan,. Expires Apr 19,2016 [Page 5] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 3 UseCase Scenarios (Simulated in Lab) and Results Data for enhancement opportunities are derived from our lab tests. These tests are done in a reasonably populated, yet contained test network. The initial set of tests are more focused on the throughput side and covers simulated Near, Mid and Far cell network conditions. The Intention is to get detailed data from set of tests to cover different types of data access (short/long or elastic/inelastic applications, mobile network conditions,..etc) as well as different mptcp profiles (for eg. number of sub flows). The detailed analysis and summary would be presented in the later sections of the document, followed by design/implementation recommendations for the SPs. 3.1 MPTCP Enabled Client Uploads data from Non-MPTCP Capable Server 3.2 MPTCP Enabled Client Uploads data from MPTCP Enabled Server 3.3 MPTCP Enabled Client Uploads data from Non-MPTCP Capable Server with Intermediate MPTCP Enabled devices (proxy?) 3.4 MPTCP Enabled Client Uploads data from MPTCP Enabled Server with Intermediate MPTCP Enabled devices (proxy?) Palanivelan,Chetan,. Expires Apr 19,2016 [Page 6] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 3.5 MPTCP Enabled Client Downloads data from Non-MPTCP Capable Server 3.6 MPTCP Enabled Client Downloads data from MPTCP Enabled Server 3.7 MPTCP Enabled Client Downloads data from Non-MPTCP Capable Server with Intermediate MPTCP Enabled devices (proxy?) 3.8 MPTCP Enabled Client Downloads data from MPTCP Enabled Server with Intermediate MPTCP Enabled devices (proxy?) Palanivelan,Chetan,. Expires Apr 19,2016 [Page 7] INTERNET DRAFT MPTCP Enhancement Opportunities Oct 17,2015 4 Security Considerations None 5 IANA Considerations None 6 References 6.1 Normative References [RFC2119] S. Bradner, "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. 6.2 Informative References [RFC6182] Ford, A., Raiciu, C., Handley, M., Barre, S., and J.Iyengar, "Architectural Guidelines for Multipath TCP Development", RFC 6182, March 2011. [RFC6356] Raiciu, C., Handley, M., and D. Wischik,"Coupled Congestion Control for Multipath Transport Protocols", RFC 6356,October 2011. Author's Addresses Palanivelan Appanasamy Manager Technology, Verizon Labs Bangalore, India Email: palanivelan.appanasamy@verizon.com Chetan Harsha R&D Software Engineer, Verizon Labs Bangalore, India Email: chetan.harsha@verizon.com Senthil sivakumar Principal Engineer, Cisco systems Durham, NC Email: ssenthil@cisco.com Palanivelan,Chetan,. Expires Apr 19,2016 [Page 8]