Internet DRAFT - draft-bryskin-use-cases-sf-aware-topo-model
draft-bryskin-use-cases-sf-aware-topo-model
Network Working Group I. Bryskin
Internet-Draft Huawei Technologies
Intended status: Informational X. Liu
Expires: December 29, 2017 Jabil
J. Guichard
Y. Lee
Huawei Technologies
June 27, 2017
Use Cases for SF Aware Topology Models
draft-bryskin-use-cases-sf-aware-topo-model-00
Abstract
This document describes some use cases that benefit from the network
topology models that are service and network function aware.
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). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at http://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 29, 2017.
Copyright Notice
Copyright (c) 2017 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
Bryskin, et al. Expires December 29, 2017 [Page 1]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Exporting SF/NF Information to Network Clients and Other
Network SDN Controllers . . . . . . . . . . . . . . . . . . . 3
3. Flat End-to-end SFCs Managed on Multi-domain Networks . . . 3
4. Managing SFCs with TE Constrains . . . . . . . . . . . . . . 4
5. SFC Protection and Load Balancing . . . . . . . . . . . . . . 5
6. Network Clock Synchronization . . . . . . . . . . . . . . . . 5
7. Client - Provider Network Slicing Interface . . . . . . . . . 6
8. Dynamic Assignment of Regenerators for L0 Services . . . . . 6
9. Dynamic Assignment of OAM Functions for L1 Services . . . . . 7
10. SFC Abstraction and Scaling . . . . . . . . . . . . . . . . . 7
11. Dynamic Compute/VM/Storage Resource Assignment . . . . . . . 8
12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8
13. Security Considerations . . . . . . . . . . . . . . . . . . . 8
14. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 8
15. References . . . . . . . . . . . . . . . . . . . . . . . . . 8
15.1. Normative References . . . . . . . . . . . . . . . . . . 8
15.2. Informative References . . . . . . . . . . . . . . . . . 9
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 9
1. Introduction
Normally network connectivity services are discussed as a means to
interconnect various abstract or physical network topological
elements, such as ports, link termination points and nodes
[I-D.ietf-teas-yang-te-topo] [I-D.ietf-teas-yang-te]. However, the
connectivity services, strictly speaking, interconnect not the
network topology elements per-se, rather, located on/associated with
the various network and service functions [RFC7498] [RFC7665]. In
many scenarios it is beneficial to decouple the service/network
functions from the network topology elements hosting them, describe
them in some unambiguous and identifiable way (so that it would be
possible, for example, to auto-discover on the network topology
service/network functions with identical or similar functionality and
characteristics) and engineer the connectivity between the service/
network functions, rather than between their current topological
locations. The purpose of this document is to describe some use
cases that could benefit from such an approach.
Bryskin, et al. Expires December 29, 2017 [Page 2]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
2. Exporting SF/NF Information to Network Clients and Other Network SDN
Controllers
In the context of Service Function Chain (SFC) orchestration one
existing problem is that there is no way to formally describe a
Service or Network Function in a standard way (recognizable/
understood by a third party) as a resource of a network topology
node.
One implication of this is that there is no way for the orchestrator
to give a network client even a ball-park idea as to which network's
SFs/NFs are available for the client's use/control and where they are
located in the network even in terms of abstract topologies/virtual
networks configured and managed specifically for the client.
Consequently the client has no say on how the SFCs provided for the
client by the network should be set up and managed (which SFs are to
be used and how they should be chained together, optimized,
manipulated, protected, etc.).
Likewise, there is no way for the orchestrator to export SF/NF
information to other network controllers. The SFC orchestrator may
serve, for example, a higher level controller (such as Network
Slicing Orchestrator), with the latter wanting at least some level of
control as to which SFs/NFs it wants on its SFCs and how the Service
Function Paths (SFPs) are to be routed and provisioned, especially,
if it uses services of more than one SFC orchestrator.
The issue of exporting of SF/NF information could be addressed by
defining a model, in which formally described/recognizable SF/NF
instances are presented as topological elements, for example, hosted
by (associated with) TE, L3 or L2 topology nodes. The model could
describe whether, how and at what costs the SFs/NFs hosted by a given
node could be chained together, how these intra-node SFCs could be
connected to the node's Service Function Forwarders (SFFs, entities
dealing with SFC NSHs and metadata), and how the SFFs could be
connected to the node's Tunnel and Link Termination Points (TTPs and
LTPs) to chain the intra-node SFCs across the network topology.
3. Flat End-to-end SFCs Managed on Multi-domain Networks
SFCs may span multiple administrative domains, each of which
controlled by a separate SFC controller. The usual solution for such
q scenario is the Hierarchical SFCs (H-SFCs), in which the higher
level orchestrator controls only SFs located on domain border nodes.
Said higher level SFs are chained together into higher level SFCs via
lower level (intra-domain) SFCs provisioned and controlled
independently by respective domain controllers. The decision as to
which higher level SFCs are connected to which lower level SFCs is
Bryskin, et al. Expires December 29, 2017 [Page 3]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
driven by packet re- classification every time the packet enters a
given domain. Said packet re-classification is very time-consuming
operation. Furthermore, the independent nature of higher and lower
level SFC control is prone to configuration errors, which may lead to
long lasting loops and congestions. It is highly desirable to be
able to set up and manage SFCs spanning multiple domains in a flat
way as far as the data plane is concerned (i.e. with a single packet
classification at the ingress into the multi-domain network but
without re-classifications on domain ingress nodes).
One way to achieve this is to have the domain controllers expose SF/
NF- aware topologies, and have the higher level orchestrator operate
on the network-wide topology, the product of merging of the
topologies catered by the domain controllers. This is similar in
spirit to setting up, coordinating and managing the transport
connectivity (TE tunnels) on a multi-domain multi-vendor transport
network.
4. Managing SFCs with TE Constrains
Some SFCs require per SFC link/element and end-to-end TE constrains
(bandwidth, delay/jitter, fate sharing/diversity. etc.). Said
constraints could be ensured via carrying SFPs inside overlays that
are traffic engineered with the constrains in mind. A good analogy
would be orchestrating delay constrained L3 VPNs. One way to support
such L3 VPNs is to carry MPLS LSPs interconnecting per-VPN VRFs
inside delay constrained TE tunnels interconnecting the PEs hosting
the VRFs.
Planning, computing and provisioning of TE overlays to constrain
arbitrary SFCs, especially those that span multiple administrative
domains with each domain controlled by a separate controller, is a
very difficult challenge. Currently it is addressed by pre-
provisioning on the network of multiple TE tunnels with various TE
characteristics, and "nailing down" SFs/NFs to "strategic" locations
(e.g. nodes terminating many of such tunnels) in a hope that an
adequate set of tunnels could be found to carry the SFP of a given
TE-constrained SFC. Such an approach is especially awkward in the
case when some or all of the SFs/NFs are VNFs (i.e. could be
instantiated at multiple network locations).
SF/NF-aware TE topology model in combination with TE tunnel model
will allow for the network orchestrator (or a client controller) to
compute, set up and manipulate the TE overlays in the form of TE
tunnel chains. Said chains could be duel-optimized compromising on
optimal SF/NF locations with optimal TE tunnels interconnecting them.
The TE tunnel chains (carrying multiple similarly constrained SFPs)
Bryskin, et al. Expires December 29, 2017 [Page 4]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
could be adequately constrained both at individual TE tunnel level
and at the chain end-to-end level.
5. SFC Protection and Load Balancing
Currently the combination of TE topology & tunnel models offers to a
network controller various capabilities to recover an individual TE
tunnel from network failures occurred on one or more network links or
transit nodes on the TE paths taken by the TE tunnel's connection(s).
However, there is no simple way to recover a TE tunnel from a failure
affecting its source or destination node. SF/NF-aware TE topology
model can decouple the association of a given SF/NF with its location
on the network topology by presenting multiple, identifiable as
mutually substitutable SFs/NFs hosted by different TE topology nodes.
So, for example, if it is detected that a given TE tunnel destination
node is malfunctioning or has gone out of service, the TE tunnel
could be re- routed to terminate on a different node hosting
functionally the same SFs/NFs as ones hosted by the failed node.
This is in line with the ACTN edge migration and function mobility
concepts. It is important to note that the described strategy works
much better for the stateless SFs/NFs. This is because getting the
alternative stateful SFs/MFs into the same respective states as the
current (i.e. active, affected by failure) are is a very difficult
challenge.
At the SFC level the SF/NF-aware TE topology model can offer SFC
dynamic restoration capabilities against failed/malfunctioning SFs/
NFs by identifying and provisioning detours to a TE tunnel chain, so
that it starts carrying the SFC's SFPs towards healthy SFs/NFs that
are functionally the same as the failed ones. Furthermore, multiple
parallel TE tunnel chains could be pre-provisioned for the purpose of
SFC load balancing and end-to-end protection. In the latter case
said parallel TE tunnel chains could be placed to be sufficiently
disjoint from each other.
6. Network Clock Synchronization
Many current and future network applications (including 5g and IoT
applications) require very accurate time services (PTP level, ns
resolution). One way to implement the adequate network clock
synchronization for such services is via describing network clocks as
NFs on an NF-aware TE topology optimized to have best possible delay
variation characteristics. Because such a topology will contain
delay/delay variation metrics of topology links and node cross-
connects, as well as costs in terms of delay/delay variation of
connecting clocks to hosting them node link and tunnel termination
points, it will be possible to dynamically select and provision bi-
directional time-constrained deterministic paths or trees connecting
Bryskin, et al. Expires December 29, 2017 [Page 5]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
clocks (e.g. grand master and boundary clocks) for the purpose of
exchange of clock synchronization information. Note that network
clock aware TE topologies separately provided by domain controllers
will enable multi-domain network orchestrator to set up and
manipulate the clock synchronization paths/trees spanning multiple
network domains.
7. Client - Provider Network Slicing Interface
3GPP defines network slice as "a set of network functions and the
resources for these network functions which are arranged and
configured, forming a complete logical network to meet certain
network characteristics.". Network slice could be also defined as a
logical partition of a provider's network that is owned and managed
by a tenant. SF/NF-aware TE topology model has a potential to
support a very important interface between network slicing clients
and providers because, on the one hand, the model can describe
holistically and hierarchically the client's requirements and
preferences with respect to a network slice functional, topological
and traffic engineering aspects, as well as of the degree of resource
separation/ sharing between the slices, thus allowing for the client
(up to agreed upon extent) to dynamically (re-)configure the slice or
(re-)schedule said (re-)configurations in time, while, on the other
hand, allowing for the provider to convey to the client the slice's
operational state information and telemetry the client has expressed
interest in.
8. Dynamic Assignment of Regenerators for L0 Services
On large optical networks some of provided to their clients L0
services could not be provisioned as single Och trails, rather, as
chains of such trails interconnected via regenerators, such as 3R
regenerators. Current practice of the provisioning of such services
requires configuration of explicit paths (EROs) describing identity
and location of regenerators to be used. A solution is highly
desirable that could:
o Identify such services based, for example, on optical impairment
computations;
o Assign adequate for the services regenerators dynamically out of
the regenerators that are grouped together in pools and
strategically scattered over the network topology nodes;
o Compute and provision supporting the services chains of optical
trails interconnected via so selected regenerators, optimizing the
chains to use minimal number of regenerators, their optimal
Bryskin, et al. Expires December 29, 2017 [Page 6]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
locations, as well as optimality of optical paths interconnecting
them;
o Ensure recovery of such chains from any failures that could happen
on links, nodes or regenerators along the chain path.
NF-aware TE topology model (in this case L1 NF-aware L0 topology
model) is just the model that could provide a network controller (or
even a client controller operating on abstract NF-aware topologies
provided by the network) to realize described above computations and
orchestrate the service provisioning and network failure recovery
operations.
9. Dynamic Assignment of OAM Functions for L1 Services
OAM functionality is normally managed by configuring and manipulating
TCM/MEP functions on network ports terminating connections or their
segments over which OAM operations, such as performance monitoring,
are required to be performed. In some layer networks (e.g.
Ethernet) said TCMs/MEPs could be configured on any network ports.
In others (e.g. OTN/ODUk) the TCMs/MEPs could be configured on some
(but not all network ports) due to the fact that the OAM
functionality (i.e. recognizing and processing of OAM messages,
supporting OAM protocols and FSMs) requires in these layer networks
certain support in the data plane, which is not available on all
network nodes. This makes TCMs/MEPs good candidates to be modeled as
NFs. This also makes TCM/MEP aware topology model a good basis for
placing dynamically an ODUk connection to pass through optimal OAM
locations without mandating the client to specify said locations
explicitly.
10. SFC Abstraction and Scaling
SF/NF-aware topology may contain information on native SFs/NFs (i.e.
SFs/NFs as known to the provider itself) and/or abstract SFs/NFs
(i.e. logical/macro SFs/NFs representing one or more SFCs each made
of native and/or lower level abstract SFs/NFs). As in the case of
abstracting topology nodes, abstracting SFs/NFs is hierarchical in
nature - the higher level of SF/NF-aware topology, the "larger"
abstract SFs/NFs are, i.e. the larger data plane SFCs they represent.
This allows for managing large scale networks with great number of
SFs/NFs (such as Data Center interconnects) in a hierarchical, highly
scalable manner resulting in control of very large number of flat in
the data plane SFCs that span multiple domains.
Bryskin, et al. Expires December 29, 2017 [Page 7]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
11. Dynamic Compute/VM/Storage Resource Assignment
In a distributed data center networks, virtual machines for compute
resources may need to be dynamically re-allocated due to various
reasons such as DCI network failure, compute resource load balancing,
etc. In many cases, the DCI connectivity for the source and the
destination is not predetermined. There may be a pool of sources and
a pool of destination data centers associated with re-allocation of
compute/VM/storage resources. There is no good mechanism to date to
capture this dynamicity nature of compute/VM/storage resource
reallocation. Generic Compute/VM/Storage resources can be described
and announced as a SF where a DC hosting these resources can be
modeled as an abstract node. Topology interconnecting these abstract
nodes (DCs) in general is of multi-domain nature. Thus, SF-aware
topology model can facilitate a joint optimization of TE network
resources and Compute/VM/Storage resources and solve Compute/VM/
Storage mobility problem within and between DCs.
12. IANA Considerations
This document has no actions for IANA.
13. Security Considerations
This document does not define networking protocols and data, hence
are not directly responsible for security risks.
14. Acknowledgements
The authors would like to thank Maarten Vissers, Joel Halpern, and
Greg Mirsky for their helpful comments and valuable contributions.
15. References
15.1. Normative References
[RFC7498] Quinn, P., Ed. and T. Nadeau, Ed., "Problem Statement for
Service Function Chaining", RFC 7498,
DOI 10.17487/RFC7498, April 2015,
<http://www.rfc-editor.org/info/rfc7498>.
[RFC7665] Halpern, J., Ed. and C. Pignataro, Ed., "Service Function
Chaining (SFC) Architecture", RFC 7665,
DOI 10.17487/RFC7665, October 2015,
<http://www.rfc-editor.org/info/rfc7665>.
Bryskin, et al. Expires December 29, 2017 [Page 8]
Internet-Draft Use Cases for SF Aware Topo Models June 2017
[I-D.ietf-teas-yang-te-topo]
Liu, X., Bryskin, I., Beeram, V., Saad, T., Shah, H., and
O. Dios, "YANG Data Model for TE Topologies", draft-ietf-
teas-yang-te-topo-09 (work in progress), June 2017.
[I-D.ietf-teas-yang-te]
Saad, T., Gandhi, R., Liu, X., Beeram, V., Shah, H., and
I. Bryskin, "A YANG Data Model for Traffic Engineering
Tunnels and Interfaces", draft-ietf-teas-yang-te-06 (work
in progress), March 2017.
15.2. Informative References
[I-D.ietf-sfc-hierarchical]
Dolson, D., Homma, S., Lopez, D., Boucadair, M., Liu, D.,
Ao, T., and V. Vu, "Hierarchical Service Function Chaining
(hSFC)", draft-ietf-sfc-hierarchical-02 (work in
progress), January 2017.
Authors' Addresses
Igor Bryskin
Huawei Technologies
EMail: Igor.Bryskin@huawei.com
Xufeng Liu
Jabil
EMail: Xufeng_Liu@jabil.com
Jim Guichard
Huawei Technologies
EMail: james.n.guichard@huawei.com
Young Lee
Huawei Technologies
EMail: leeyoung@huawei.com
Bryskin, et al. Expires December 29, 2017 [Page 9]