| TEAS Working Group | A. Wang |
| Internet-Draft | China Telecom |
| Intended status: Experimental | Q. Zhao |
| Expires: April 24, 2019 | B. Khasanov |
| H. Chen | |
| Huawei Technologies | |
| R. Mallya | |
| Juniper Networks | |
| October 21, 2018 |
PCE in Native IP Network
draft-ietf-teas-pce-native-ip-02
This document defines the CCDR framework for traffic engineering within native IP network, using Dual/Multi-BGP session strategy and PCE-based central control architecture. The proposed central mode control framework conforms to the concept that defined in [RFC8283]. The scenario and simulation results of CCDR traffic engineering is described in draft [I-D.ietf-teas-native-ip-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 April 24, 2019.
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.
Draft [I-D.ietf-teas-native-ip-scenarios] describes the scenario and simulation results for traffic engineering in native IP network. In summary, the requirements for traffic engineering in native IP network are the followings:
[I-D.ietf-pce-pcep-extension-native-ip].
This document defines the framework for traffic engineering within native IP network, using Dual/Multi-BGP session strategy, to meet the above requirements in dynamical and central control mode. The related PCEP protocol extensions to transfer the key parameters between PCE and the underlying network devices(PCC) are provided in draft
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] .
Dual-BGP framework for simple topology is illustrated in Fig.1, which is comprised by SW1, SW2, R1, R2. There are multiple physical links between R1 and R2. Traffic between IP11 and IP21 is normal traffic, traffic between IP12 and IP22 is priority traffic that should be treated differently.
Only native IGP/BGP protocol is deployed between R1 and R2. The traffic between each address pair may change timely and the corresponding source/destination addresses of the traffic may also change dynamically.
The key ideas of the Dual-BGP framework for this simple topology are the followings:
The traffic between the IP11 and IP21, and the traffic between IP12 and IP22 will go through different physical links between R1 and R2, each type of traffic occupy different dedicated physical links.
If there is more traffic between IP12 and IP22 that needs to be assured , one can add more physical links between R1 and R2 to reach the loopback address lo1(also the next hop for BGP Peer pair2). In this cases the prefixes that advertised by two BGP peers need not be changed.
If, for example, there is traffic from another address pair that needs to be assured (for example IP13/IP23), and the total volume of assured traffic does not exceed the capacity of the previous appointed physical links, one need only to advertise the newly added source/destination prefixes via the BGP peer pair2. The traffic between IP13/IP23 will go through the assigned dedicated physical links as the traffic between IP12/IP22.
Such decouple philosophy gives network operator flexible control ability on the network traffic, achieve the determined QoS assurance effect to meet the application's requirement. No complex MPLS signal procedures is introduced, the router need only support native IP protocol.
| BGP Peer Pair2 |
+------------------+
|lo1 lo1 |
| |
| BGP Peer Pair1 |
+------------------+
IP12 |lo0 lo0 | IP22
IP11 | | IP21
SW1-------R1-----------------R2-------SW2
Links Group
Fig.1 Design Philosophy for Dual-BGP Framework
When the assured traffic spans across one large scale network, as that illustrated in Fig.2, the dual BGP sessions cannot be established hop by hop especially for the iBGP within one AS.
For such scenario, we should consider to use the Route Reflector (RR) to achieve the similar Dual-BGP effect, select one router which performs the role of RR (for example R3 in Fig.2), every other edge router will establish two BGP peer sessions with the RR, using their different loopback addresses respectively. The other two steps for traffic differentiation are same as that described in the Dual-BGP simple topology usage case.
For the example shown in Fig.2, if we select the R1-R2-R4-R7 as the dedicated path, then we should set the explicit peer routes on these routers respectively, pointing to the BGP next hop (loopback addresses of R1 and R7, which are used to send the prefix of the assured traffic) to the actual address of the physical link.
+------------R3--------------+
| |
SW1-------R1-------R5---------R6-------R7--------SW2
| | | |
+-------R2---------R4--------+
Fig.2 Dual-BGP Framework for Large Scale Network
In general situation, several additional traffic differentiation criteria exist, including:
These different traffic requirements can be summarized in the following table:
+----------+-------------+---------------+-----------------+
| Flow No. | Latency | Packet Loss | Jitter |
+----------+-------------+---------------+-----------------+
| 1 | Low | Normal | Don't care |
+----------+-------------+---------------+-----------------+
| 2 | Normal | Low | Dont't care |
+----------+-------------+---------------+-----------------+
| 3 | Normal | Normal | Low |
+----------+-------------+---------------+-----------------+
Table 1. Traffic Requirement Criteria
It is almost impossible to provide an end-to-end (E2E) path with latency, jitter, packet loss constraints to meet the above requirements in large scale IP-based network via the distributed routing protocol, but these requirements can be solved using the CCDR framework since the PCE has the overall network view, can collect real network topology and network performance information about the underlying network, select the appropriate path to meet various network performance requirements of different traffic.
The procedures to implement the Multi-BGP strategy are the followings:
+----+
***********+ PCE+*************
* +--*-+ *
* / * \ *
* * *
PCEP* BGP-LS/SNMP *PCEP
* * *
* * \ * /
\ * / * \ */
\*/-----------R3--------------*
| |
| |
SW1-------R1-------R5---------R6-------R7--------SW2
| | | |
| | | |
+-------R2---------R4--------+
Fig.3 PCE based framework for Multi-BGP deployment
The PCEP protocol needs to be extended to transfer the following key parameters:
[I-D.ietf-pce-pcep-extension-native-ip] .
Once the router receives such information, it should establish the BGP session with the peer appointed in the PCEP message, advertise the prefixes that contained in the corresponding PCEP message, and build the end to end dedicated path hop by hop. Details of communications between PCEP and BGP subsystems in router's control plane are out of scope of this draft and will be described in separate draft
The reason that we selected PCEP as the southbound protocol instead of OpenFlow, is that PCEP is suitable for the changes in control plane of the network devices, there OpenFlow dramatically changes the forwarding plane. We also think that the level of centralization that requires by OpenFlow is hardly achievable in many today’s SP networks so hybrid BGP+PCEP approach looks much more interesting.
In CCDR framework, PCE needs only to influence the edge routers for the prefixes differentiation via the multi-BGP deployment. The route information for these prefixes within the on-path routers were distributed via the BGP protocol. Unlike the solution from BGP Flowspec, the on-path router need only keep the specific policy routes to the BGP next-hop of the differentiate prefixes, not the specific routes to the prefixes themselves. This can lessen the burden from the table size of policy based routes for the on-path routers, and has more scalabilities when comparing with the solution from BGP flowspec or Openflow.
CCDR framework is based on the distributed IP protocol. If the PCE failed, the forwarding plane will not be impacted, as the BGP session between all devices will not flap, and the forwarding table will remain the same. If one node on the optimal path is failed, the assurance traffic will fall over to the best-effort forwarding path. One can even design several assurance paths to load balance/hot standby the assurance traffic to meet the path failure situation, as done in MPLS FRR.
For high availability of PCE/SDN-controller, operator should rely on existing HA solutions for SDN controller, such as clustering technology and deployment.
Not every router within the network support will support the PCEP extension that defined in [I-D.ietf-pce-pcep-extension-native-ip] simultaneously. For such situations, router on the edge of domain can be upgraded first, and then the traffic can be assured between different domains. Within each domain, the traffic will be forwarded along the best-effort path. Service provider can selectively upgrade the routers on each domain in sequence.
Solution described in this draft puts more requirements on the function of PCE and its communication with the underlay devices. The PCE should have the capability to calculate the loop-free e2e path upon the status of network condition and the service requirements in real time. The PCE need also to consider the router order during deployment to eliminate the possible transient traffic loop.
This solution does not require the change of forward behavior on the underlay devices, then there will no additional security impact for the devices.
When deploy the solution on network, service provider should also consider more on the protection of SDN controller and their communication with the underlay devices, which is described in document [RFC5440] and [RFC8253]
This document does not require any IANA actions.
Penghui Mi and Shaofu Peng contribute the contents of this draft.
The author would like to thank Deborah Brungard, Adrian Farrel, Huaimo Chen, Vishnu Beeram, Lou Berger, Dhruv Dhody and Jessica Chen for their supports and comments on this draft.