Internet DRAFT - draft-liu-rtgwg-mdt-in-high-bdp
draft-liu-rtgwg-mdt-in-high-bdp
Routing Area Working Group Y. Liu
Internet Draft M. Han
Intended status: Informational Track China Unicom
Expires: September 4, 2024 March 4, 2024
Use Cases and Requirements of Massive Data Transmission（MDT）
in High Bandwidth-delay Product (BDP) Network
draft-liu-rtgwg-mdt-in-high-bdp-00
Abstract
This document describes the use cases and related requirements of
Massive Data Transmission（MDT）in High Bandwidth-delay Product (BDP)
Network. To achieve MDT, it is necessary to implement service
identification and traffic record, network layer load balancing,
transmission protocol optimization, etc.
Status of this Memo
This Internet-Draft is submitted 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 September 4, 2024.
Liu, et al. Expire September 4, 2024 [Page 1]
Internet-Draft MDT in High BDP Network March 2024
Copyright Notice
Copyright (c) 2023 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.
Table of Contents
1. Introduction..................................................2
1.1. Requirements Language.......................................3
2. Massive data transmission in high BDP network.................3
3. Use Cases and Requirements....................................4
3.1. Service Identification and Traffic Record...................4
3.2. Load Balancing at Network Level.............................5
3.3. Optimization of Transmission Protocols......................5
3.4. Collaboration Requirements..................................5
4. Security Considerations.......................................6
5. IANA Considerations...........................................6
6. References....................................................6
6.1. Normative References........................................6
6.2. Informational References....................................6
Authors' Addresses...............................................7
1. Introduction
With the continuous development of industries such as autonomous
driving, AI intelligent computing, and enterprise cloud, the demand
for massive data transmission across wide area networks from edge
data centers/enterprises to core data centers has become
increasingly common, and higher requirements have been put forward
for existing carrier network architectures.
Taking the scenarios of supercomputing and intelligent computing as
example, data transmission usually includes two requirements:
1) The transmission of training data between intelligent computing
centers, supercomputing centers, and between intelligent computing
centers and supercomputing centers is usually carried by optical
Liu, et al. Expire September 4, 2024 [Page 2]
Internet-Draft MDT in High BDP Network March 2024
networks due to high bandwidth requirements and high connection
stability.
2) The transmission of training data and result feedback between
users and intelligent computing centers/supercomputing centers can
be carried through IP networks due to their strong suddenness and
cost sensitivity.
The MDT can be achieved by traditional high-speed private lines,
providing users with efficient and reliable data transmission.
However, traditional private lines usually use billing methods such
as daily or monthly rent, with fixed bandwidth resources and
expensive prices. The long-distance transmission of massive data
requires flexible transmission tasks based on user data
characteristics, completion time, and security requirements,
utilizing the idle bandwidth resources of existing private lines and
networks to reduce transmission costs and improve transmission
efficiency.
This draft mainly describes the overall architecture of feasible
solutions for MDT in high BDP network, typical problems that may be
encountered, and proposes potential solutions, including but not
limited to how to perform load balancing scheduling at the global
level of the network to avoid the impact of massive data
transmission on existing network services; how to identify MDT
services for traffic record and billing purposes; how to
optimize the congestion control algorithm of the transport layer
protocol to ensure that the throughput of TCP protocol can be
improved in long-distance lossy networks.
1.1. Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in
BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
capitals, as shown here.
2. Massive data transmission in high BDP network
Figure 1 show schematic diagram of the network architecture of
MDT in high BDP network, where key functional units include:
DC/User Application(APP)：APP can be deployed on DC/personal terminal
devices, which can be traditional file transfer tools or customized
APP developed for MDT scenarios, which implements enhanced
functions such as intelligent data compression, intelligent
partitioning, encryption, etc.
Liu, et al. Expire September 4, 2024 [Page 3]
Internet-Draft MDT in High BDP Network March 2024
Network：existing service carrier networks of current operators, such
as metropolitan area networks, backbone networks, etc.
Controller：existing network management system includes controllers,
collaborators, orchestrators, etc.
+-----+ +-----------+ +------+
| DC |<--------------------------->| | |Center|
| APP | |---->|Controller |<-----| | DC2 |
+-----+ | +-----------| | |->+------+
| | ^ | |
| +------+ | | | |
| | Edge | V V V |
|--->| DC1 |-\ +-------+ +-------+ +-------+ /
+------+ \-> | Edge | | Core | | Edge |/
| |-->| |-->| |
/------> |Network| |Network| |Network|
/ +-------+ +-------+ +-------+\
+------+ / \ +------+
| User1|--/ \-> | User2|
| APP | | APP |
+------+ +------+
Figure 1: Architecture of MDT in high BDP network
3. Use Cases and Requirements
MDT service is a predictable time-efficient service that requires
data transmission to be completed within a specified time, not
sensitive to transmission delay, but requires a considerable amount
of network resources. Compared with traditional Internet and private
line services, how to improve transmission efficiency, achieve
service identification, complete scheduling and billing for services
are key issues to be considered.
3.1. Service Identification and Traffic Record
Before starting transmission, the APP will notify the controller of
the required data size and expected completion time for the task.
The controller will dynamically adjust the network path calculation
and private line bandwidth based on transmission requirements and
the current available link resources of the network, and distribute
the configuration to network nodes.
After the transmission task is initiated, network devices need to be
able to identify MDT services and corresponding account information
based on certain identifiers, perform traffic record, and report
the statistical results to the controller. The controller can get
Liu, et al. Expire September 4, 2024 [Page 4]
Internet-Draft MDT in High BDP Network March 2024
the overall MDT usage of the network, as well as the specific
resource completion status for a particular user, and make
corresponding strategy adjustments.
From the above use cases, it can be seen that the billing of MDT
services and the scheduling and allocation of available bandwidth
resources in the current network require network devices to
recognize the current MDT services. APNID defined in [I-D.li-apn-
problem-statement-usecases] and [I-D.li-apn-header] may be a
potential solution to meet the identification requirements of MDT
service.
3.2. Load Balancing at Network Level
The bandwidth requirement for MDT service is generally between 500M-
10G, and the launch of each service requires a huge consumption of
network resources. With the continuous increase of service
launching, how to make reasonable use of network idle resources,
allocate global network resources and data express tasks, and
minimize the impact on existing services have become necessary
issues to consider.
The controller notifies the APP of available network resource
information, and the APP dynamically adjusts the data sending
strategy based on the available network bandwidth, and cooperates
with network devices to improve the overall resource utilization of
the network. When the controller discovers a shortage of available
network resources or predicts a rapid growth in future network
traffic, it should notify the APP side in advance to make policy
adjustments.
3.3. Optimization of Transmission Protocols
In most scenarios, the two ends of MDT services need to cross a wide
area network, with a distance of over 1000 km. RTT is in the tens of
MS range, and there is a small amount of packet loss in the network,
which poses new challenges to the traditional TCP [RFC7805].
Based on current test results, the traditional TCP congestion
control algorithm [RFC2581] may not achieve the expected
transmission rate for MDT. Therefore, an efficient, secure
transmission protocol that can adapt to the current network state
and resource status is needed to solve these problems.
3.4. Collaboration Requirements
TBD
Liu, et al. Expire September 4, 2024 [Page 5]
Internet-Draft MDT in High BDP Network March 2024
4. Security Considerations
TBD.
5. IANA Considerations
TBD.
6. References
6.1. Normative References
[RFC7805] Zimmermann, A., Eddy, W., and L. Eggert, "Moving Outdated
TCP Extensions and TCP-Related Documents to Historic or
Informational Status", RFC 7805, DOI 10.17487/RFC7805,
April 2016, <https://www.rfc-editor.org/info/rfc7805>.
[RFC2581] Allman, M., Paxson, V., and W. Stevens, "TCP Congestion
Control", RFC 2581, DOI 10.17487/RFC2581, April 1999,
<https://www.rfc-editor.org/info/rfc2581>.
[RFC 9526] Filsfils, C., Talaulikar, K., Ed., Voyer, D., Bogdanov,
A., and P. Mattes, "Segment Routing Policy Architecture",
RFC 9256, DOI 10.17487/RFC9256, July 2022,
<https://www.rfc-editor.org/info/rfc9256>.
6.2. Informational References
[I-D.li-apn-problem-statement-usecases] Li, Z., Peng, S., Voyer, D.,
Xie, C., Liu, P., Qin, Z., and G. S. Mishra, "Problem
Statement and Use Cases of Application-aware Networking
(APN)", Work in Progress, Internet-Draft, draft-li-apn-
problem-statement-usecases-08, 3 April 2023,
<https://datatracker.ietf.org/doc/html/draft-li-apn-
problem-statement-usecases-08>.
[I-D.li-apn-header] Li, Z., Peng, S., and S. Zhang, "Application-
aware Networking (APN) Header", Work in Progress,
Internet-Draft, draft-li-apn-header-04, 12 April 2023,
<https://datatracker.ietf.org/doc/html/draft-li-apn-
header-04>.
Liu, et al. Expire September 4, 2024 [Page 6]
Internet-Draft MDT in High BDP Network March 2024
Authors' Addresses
Ying Liu
China Unicom
China
Email: liuy619@chinaunicom.cn
Mengyao Han
China Unicom
China
Email: hanmy12@chinaunicom.cn
Zheng Ruan
China Unicom
China
Email: ruanz6@chinaunicom.cn
Liu, et al. Expire September 4, 2024 [Page 7]
