| TEAS Working Group | A. Wang |
| Internet-Draft | China Telecom |
| Intended status: Experimental | X. Huang |
| Expires: December 28, 2018 | C. Kou |
| BUPT | |
| Z. Li | |
| China Mobile | |
| L. Huang | |
| P. Mi | |
| Huawei Technologies | |
| June 26, 2018 |
CCDR Scenario, Simulation and Suggestion
draft-ietf-teas-native-ip-scenarios-01
This document describes the scenarios, simulation and suggestions for the “Centrally Control Dynamic Routing (CCDR)” architecture, which integrates the merit of traditional distributed protocols (IGP/BGP), and the power of centrally control technologies (PCE/SDN) to provide one feasible traffic engineering solution in various complex scenarios for the service provider.
Traditional MPLS-TE solution is mainly used in static network planning scenario and is difficult to meet the QoS assurance requirements in real-time traffic network. With the emerge of SDN concept and related technologies, it is possible to simplify the complexity of distributed control protocol, utilize the global view of network condition, give more efficient solution for traffic engineering in various complex scenarios.
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). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on December 28, 2018.
Copyright (c) 2018 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 (https://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.
Internet network is composed mainly tens of thousands of routers that run distributed protocol to exchange the reachability information between them. The path for the destination network is mainly calculated and controlled by the traditional IGP protocols. These distributed protocols are robust enough to support the current evolution of Internet but has some difficulties when the application requires the end-to-end QoS performance, or the service provider wants to maximize the links utilization within their network.
MPLS-TE technology is one perfect solution for the finely planned network but it will put heavy burden on the router when we use it to solve the dynamic QoS assurance requirements within real time traffic network.
SR(Segment Routing) is another prominent solution that integrates some merits of traditional distributed protocol and the advantages of centrally control mode, but it requires the underlying network, especially the provider edge router to do label push and pop action in-depth, and need some complex solutions for co-exist with the Non-SR network. Finally, it can only maneuver the end-to-end path for MPLS and IPv6 traffic via different mechanisms.
The advantage of MPLS is mainly for traffic isolation, such as the L2/L3 VPN service deployments, but most of the current application requirements are only for high performances end-to-end QoS assurance. Without the help of centrally control architecture, the service provider almost can’t make such SLA guarantees upon the real time traffic situation.
This draft gives some scenarios that the centrally control dynamic routing (CCDR) architecture can easily solve, without adding more extra burdening on the router. It also gives the PCE algorithm results under the similar topology, traffic pattern and network size to illustrate the applicability of CCDR architecture. Finally, it gives some suggestions for the implementation and deployment of CCDR.
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].
The following sections describe some scenarios that the CCDR architecture is suitable for deployment.
With the emerge of cloud computing technologies, enterprises are putting more and more services on the public oriented service infrastructure, but keep still some core services within their network. The bandwidth requirements between the private cloud and the public cloud are occasionally and the background traffic between these two sites varied from time to time. Enterprise cloud applications just want to invoke the network capabilities to make the end-to-end QoS assurance on demand. Otherwise, the traffic should be controlled by the distributed protocol.
CCDR, which integrates the merits of distributed protocol and the power of centrally control, is suitable for this scenario. The possible solution architecture is illustrated below:
+------------------------+
| Cloud Based Application|
+------------------------+
|
+-----------+
| PCE |
+-----------+
|
|
//--------------\\
///// \\\\\
Private Cloud Site || Distributed |Public Cloud Site
| Control Network |
\\\\\ /////
\\--------------//
Fig.1 Hybrid Cloud Communication Scenario
By default, the traffic path between the private cloud site and public cloud site will be determined by the distributed control network. When some applications require the end-to-end QoS assurance, it can send these requirements to PCE, let PCE compute one e2e path which is based on the underlying network topology and the real traffic information, to accommodate the application’s QoS requirements. The proposed solution can refer the draft [I-D.ietf-teas-pce-native-ip]. Section 4 describes the detail simulation process and the results.
Currently, the network topology within MAN is generally in star mode as illustrated in Fig.2, with the different devices connect different customer types. The traffic pattern of these customers demonstrates some tidal phenomena that the links between the CR/BRAS and CR/SR will experience congestion in different periods because the subscribers under BRAS often use the network at night and the dedicated line users under SR often use the network during the daytime. The uplink between BRAS/SR and CR must satisfy the maximum traffic pattern between them and this causes the links underutilization.
+--------+
| CR |
+----|---+
|
--------|--------|-------|
| | | |
+--|-+ +-|- +--|-+ +-|+
|BRAS| |SR| |BRAS| |SR|
+----+ +--+ +----+ +--+
Fig.2 STAR-style network topology within MAN
If we can consider link the BRAS/SR with local loop, and control the MAN with the CCDR architecture, we can exploit the tidal phenomena between BRAS/CR and SR/CR links, increase the efficiency of them.
+-------+
----- PCE |
| +-------+
+----|---+
| CR |
+----|---+
|
--------|--------|-------|
| | | |
+--|-+ +-|- +--|-+ +-|+
|BRAS-----SR| |BRAS-----SR|
+----+ +--+ +----+ +--+
Fig.3 Increase the link utilization via CCDR
The operator’s networks are often comprised by tens of different domains, interconnected with each other, form very complex topology that illustrated in Fig.4. Due to the traffic pattern to/from MAN and IDC, the links between them are often in asymmetric style. It is almost impossible to balance the utilization of these links via the distributed protocol, but this unbalance phenomenon can be overcome via the CCDR architecture.
+---+ +---+
|MAN|-----------------IDC|
+-|-| | +-|-+
| ---------| |
------|BackBone|------
| ----|----| |
| | |
+-|-- | ----+
|IDC|----------------|MAN|
+---| |---+
Fig.4 TE within Complex Multi-Domain topology
In more general situation, there are often temporal congestion periods within part of the service provider’s network. Such congestion phenomena will appear repeatedly and if the service provider has some methods to mitigate it, it will certainly increase the satisfaction degree of their customer. CCDR is also suitable for such scenario that the traditional distributed protocol will process most of the traffic forwarding and the controller will schedule some traffic out of the congestion links to lower the utilization of them. Section 4 describes the simulation process and results about such scenario.
The following sections describe the topology, traffic matrix, end-to-end path optimization and congestion elimination in CCDR simulation.
The network topology mainly contains nodes and links information. Nodes used in simulation have two types: core nodes and edge nodes. The core nodes are fully linked to each other. The edge nodes are connected only with some of the core nodes. Fig.5 is a topology example of 4 core nodes and 5 edge nodes. In CCDR simulation, 100 core nodes and 400 edge nodes are generated.
+----+
/|Edge|\
| +----+ |
| |
| |
+----+ +----+ +----+
|Edge|----|Core|-----|Core|---------+
+----+ +----+ +----+ |
/ | \ / | |
+----+ | \ / | |
|Edge| | X | |
+----+ | / \ | |
\ | / \ | |
+----+ +----+ +----+ |
|Edge|----|Core|-----|Core| |
+----+ +----+ +----+ |
| | |
| +------\ +----+
| ---|Edge|
+-----------------/ +----+
Fig.5 Topology of simulation
The traffic matrix is generated based on the link capacity of topology. It can result in many kinds of situations, such as congestion, mild congestion and non-congestion.
In CCDR simulation, the traffic matrix is 500*500. About 20% links are overloaded when the Open Shortest Path First (OSPF) protocol is used in the network.
The CCDR end-to-end path optimization is to find the best end-to-end path which is the lowest in metric value and each link of the path is far below link’s threshold. Based on the current state of the network, PCE within CCDR architecture combines the shortest path algorithm with penalty theory of classical optimization and graph theory.
Given background traffic matrix which is unscheduled, when a set of new flows comes into the network, the end-to-end path optimization finds the optimal paths for them. The selected paths bring the least congestion degree to the network.
The link utilization increment degree(UID) when the new flows are added into the network is shown in Fig.6. The first graph in Fig.6 is the UID with OSPF and the second graph is the UID with CCDR end-to-end path optimization. The average UID of graph one is more than 30%. After path optimization, the average UID is less than 5%. The results show that the CCDR end-to-end path optimization has an eye-catching decreasing in UID relative to the path chosen based on OSPF.
+-----------------------------------------------------------+
| * * * *|
60| * * * * * *|
|* * ** * * * * * ** * * * * **|
|* * ** * * ** *** ** * * ** * * * ** * * *** **|
|* * * ** * ** ** *** *** ** **** ** *** **** ** *** **|
40|* * * ***** ** *** *** *** ** **** ** *** ***** ****** **|
UID(%)|* * ******* ** *** *** ******* **** ** *** ***** *********|
|*** ******* ** **** *********** *********** ***************|
|******************* *********** *********** ***************|
20|******************* ***************************************|
|******************* ***************************************|
|***********************************************************|
|***********************************************************|
0+-----------------------------------------------------------+
0 100 200 300 400 500 600 700 800 900 1000
+-----------------------------------------------------------+
| |
60| |
| |
| |
| |
40| |
UID(%)| |
| |
| |
20| |
| *|
| * *|
| * * * * * ** * *|
0+-----------------------------------------------------------+
0 100 200 300 400 500 600 700 800 900 1000
Flow Number
Fig.6 Simulation result with congestion elimination
Different degree of network congestion is simulated. The congestion degree (CD) is defined as the link utilization beyond its threshold.
The CCDR congestion elimination performance is shown in Fig.7. The first graph is the congestion degree before the process of congestion elimination. The average CD of all congested links is more than 10%. The second graph shown in Fig.7 is the congestion degree after congestion elimination process. It shows only 12 links among totally 20000 links exceed the threshold, and all the congestion degree is less than 3%. Thus, after schedule of the traffic in congestion paths, the degree of network congestion is greatly eliminated and the network utilization is in balance.
Before congestion elimination
+-----------------------------------------------------------+
| * ** * ** ** *|
20| * * **** * ** ** *|
|* * ** * ** ** **** * ***** *********|
|* * * * * **** ****** * ** *** **********************|
15|* * * ** * ** **** ********* *****************************|
|* * ****** ******* ********* *****************************|
CD(%) |* ********* ******* ***************************************|
10|* ********* ***********************************************|
|*********** ***********************************************|
|***********************************************************|
5|***********************************************************|
|***********************************************************|
|***********************************************************|
0+-----------------------------------------------------------+
0 0.5 1 1.5 2
After congestion elimination
+-----------------------------------------------------------+
| |
20| |
| |
| |
15| |
| |
CD(%) | |
10| |
| |
| |
5 | |
| |
| * ** * * * ** * ** * |
0 +-----------------------------------------------------------+
0 0.5 1 1.5 2
Link Number(*10000)
Fig.7 Simulation result with congestion elimination
With the above CCDR scenarios and simulation results, we can know it is necessary and feasible to find one general solution to cope with various complex situations for the most complex optimal path computation in centrally manner based on the underlay network topology and the real time traffic.
[I-D.ietf-teas-pce-native-ip] gives the principle solution for above scenarios, such thoughts can be extended to cover requirements that are more concretes in future.
This document considers mainly the integration of traditional distributed protocol and the global view of central control. It certainly can ease the management of network in various traffic-engineering scenarios described in this document, but the central control manner may also bring the new point be easily attacked. Solutions for CCDR scenarios should keep these in mind and consider more for the protection of SDN controller and their communication with the underlay devices, which described in document 1 and [RFC8253]
This document does not require any IANA actions.