Network Working Group Kohei Shiomoto (NTT) Internet Draft Dimitri Papadimitriou (Alcatel) Expires: August 2005 Jean-Louis Le Roux (France Telecom) Martin Vigoureux (Alcatel) Deborah Brungard (AT&T) February 2005 Requirements for GMPLS-based multi-region networks (MRN) draft-shiomoto-ccamp-gmpls-mrn-reqs-01.txt Status of this Memo This document is an Internet-Draft and is subject to all provisions of section 3 of RFC 3667. By submitting this Internet-Draft, each author represents that any applicable patent or other IPR claims of which he or she is aware have been or will be disclosed, and any of which he or she become aware will be disclosed, in accordance with RFC 3668. 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. Copyright Notice Copyright (C) The Internet Society (2005). All Rights Reserved. Abstract Most of the initial efforts on Generalized MPLS (GMPLS) have been related to environments hosting devices with a single switching capability, that is, one data plane switching layer. The complexity raised by the control of such data planes is similar to that seen in classical IP/MPLS networks. By extending MPLS to support multiple switching technologies, GMPLS provides a comprehensive framework for the control of a network where different types of switching capabilities coexist, which we call multi-region networks (MRN). This draft defines a framework for GMPLS based multi-region networks and lists a set of functional requirements. 1. Introduction Generalized MPLS (GMPLS) extends MPLS to handle multiple switching technologies: packet switching, layer-two switching, TDM switching, wavelength switching, and fiber switching (see [GMPLS-ARCH]). The Interface Switching Capability concept is Shiomoto et al. Expires August 2005 1 Requirements for GMPLS-based multi-region network Feb. 2005 introduced for these switching technologies and is designated as follows: PSC (packet switch capable), L2SC (Layer-2 switch capable), TDM (Time Division Multiplex capable), LSC (lambda switch capable), and FSC (fiber switch capable). Service providers may operate networks where multiple different switching capabilities exist. These networks consist of several switching technology domains, each of which uses the same switching capability. The representation, in a GMPLS control plane, of a switching technology domain is referred to as a region [HIER]. A network comprising of multiple switching capabilities, controlled by a single GMPLS control plane instance is called a Multi-Region Network (MRN). MRNs can be categorized according to the distribution of the switching capabilities amongst the LSRs: - Network elements are single switching capable LSRs and different types of LSRs form the network. All TE links terminating on such nodes have the same interface switching capability. A typical example is a network composed of PSC and TDM LSRs with only PSC TE-links and with only TDM TE-links, respectively. - Network elements are multi-switching capable LSRs i.e. nodes hosting at least more than one switching capability. TE links terminating on such nodes may have a set of one or more interface switching capabilities. A typical example is a network composed of LSRs that are capable of switching with PSC+TDM TE-links. Multi-switching capable LSRs are further classified as "simplex" and "hybrid" nodes (see Section 4.2). - Any combination of the above two elements. A network composed of both single and multi-switching capable LSRs. Since GMPLS provides a comprehensive framework for the control of different switching capabilities, a single GMPLS instance may be used to control the MRNs enabling rapid service provisioning and efficient resource usage across all switching capabilities. In GMPLS-based multi-region networks, TE Links are consolidated into a single Traffic Engineering Database (TED). Since this TED contains the information relative to all the different regions existing in the network, a path across multiple regions can be computed using this TED. Thus optimization of network resources can be sought and take place in across multiple regions. Consider, for example, a network consisting of IP/MPLS routers and TDM cross-connects. Assume that a packet LSP is routed between source and destination IP/MPLS routers, and that the LSP can be routed across the PSC-region (i.e., utilizing only resources of the IP/MPLS level topology). If the performance objective for the LSP is not satisfied, new data links may be created between the IP/MPLS routers across the TDM-region and the LSP can be routed over those links. Further, even if the LSP can be successfully established across the PSC-region, TDM FA- LSPs across the TDM region between the IP/MPLS routers may be established and used if doing so enables meeting an operatorĘs objectives on network resources available (e.g., link bandwidth, and adaptation port between regions) across the multiple regions. A service provider's network may be divided into different network layers. The customer's network is considered the highest layer network, and interfaces to the highest layer of the service provider's network. Network layers are commonly arranged according to the switching capabilities of the devices in the networks. Thus a customer network may be provided on top of the GMPLS-based multi-region network. Such customer networks may Shiomoto et al Expires August 2005 2 Requirements for GMPLS-based multi-region network Feb. 2005 take various kind of network layer. Services provided on top of GMPLS-based multi-region network is refereed to as "Multi-region network services". For example legacy IP and MPLS/IP networks can be supported on top of the multi-region networks. Details concerning requirements for such services and functionality required from multi-region networks to deliver such services will be addressed in a future release of this document. It has however to be emphasized that delivery of such services is a strong motivator for the deployment of multi-region networks. This document describes the requirements for the multi-region network. The rest of this document is organized as follows. In Section 3, the region and layer terminology considerations are provided. In Section 4, the key concepts for the Generalized MPLS-based multi-region and multi-layer service networks are described. In Section 5, the functional requirements are listed. There is no intention to specify solution specific elements in this document. The applicability of existing GMPLS protocol to MRN, and any protocol extensions, will be addressed in separate documents. 2. Conventions used in this document 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]. 3. Positioning 3.1. LSP Region and layer From the control plane viewpoint, an LSP region is defined as a set of one or several data plane layers that share the same type of switching technology. Examples of regions are: PSC, L2SC, TDM, LSC, and FSC regions. Hence, an LSP region is a technology domain (identified by the Switching Capability) for which data plane resources (i.e. data links) are represented into the control plane as an aggregate of TE information associated to a set of links (i.e. TE links). Example: VC-11 to VC4-64c capable TE links are part of the same TDM Region. On the other hand, a data plane layer is a network resource of a certain topological type (using the same type of termination functions, e.g. a VC-11 and a VC-4-64c represent two different layers), that could be used for establishing LSPs or connectionless traffic delivery. Note also that region is a control plane only concept. That is, layers of the same region share the same switching technology and, therefore, need the same set of technology specific signaling objects. Multiple layers can exist in a single region network. Moreover, the control plane mechanisms related to LSP region, e.g., Forwarding Adjacency (FA) and Virtual FA Topology, described as part of this document can also be described from a data plane multi-layer perspective. A service provider's network may be divided into different service network layers. The customer's network is considered the highest layer network, and interfaces to the highest layer of the service provider's network. Connectivity across the highest layer of the service provider's network may be provided with support from networks of successively lower layers. Network layers are commonly arranged according to the switching capabilities of the devices in the networks so that, for example, Shiomoto et al Expires August 2005 3 Requirements for GMPLS-based multi-region network Feb. 2005 there may be layer one networks (TDM, LSC and FSC) supporting layer two networks (L2SC) supporting layer three networks (IP and MPLS). The supported data plane relationship is, however, a data-plane client-server relationship where the lower layer provides a service for the higher layer using the data links of the lower layer, and so the layering relationship is actually administrative rather than dependent on the switching capabilities of the networks. Note that a multi-region network does not impact the arbitrary data plane layering of networks. 4. Key mechanisms in GMPLS-based multi-region networks An example of Multi-Region Networks (MRN) which consists of PSC and LSC is illustrated in Figure 1. The concept of region is by nature hierarchical. PSC and LSC are defined from the upper to the lower regions in Figure 1. Network elements with different switching technologies in the MRN are controlled by a unified GMPLS control plane. +-----+ | PSC | ----------| |--------- | | LSC | | | +-----+ | | | | +-----+ +-----+ +-----+ | PSC | | | | | | |-------| LSC |------| PSC | | LSC | | | | | +-----+ +-----+ +-----+ | | | | +-----+ | | | PSC | | ----------| |--------- | LSC | +-----+ Figure 1: Example of multi-region network 4.1. Interface switching capability The Interface Switching Capability (ISC) concept is introduced in GMPLS to support various kinds of switching technology in a unified way [GMPLS-ROUTING]. An ISC refers to the ability of a node to forward data of a particular type. PSC, L2SC, TDM, LSC, and FSC are defined. Each end of the link in a GMPLS network is associated with at least one switching capability. For example, PSC is associated with an interface which can delineate IP/MPLS packets (e.g., a router's interface) while LSC is associated with an interface which can switch individual wavelengths multiplexed in a fiber link (e.g., an OXC's interface). Links in the TE database are identified by their switching capabilities (at both ends). An interface may have multiple interface switching capabilities. A router has only interfaces with a single switching capability (PSC) while a hybrid node has a mixture of interfaces with single and multiple switching capabilities. 4.2. Multiple Switching Capabilities In MRN, network elements may be single-switching or multiple switching capable nodes. Single switching capable nodes will advertise a unique switching capability value as part of their Interface Switching Capability Descriptor (ISCD) sub-TLV(s) to Shiomoto et al Expires August 2005 4 Requirements for GMPLS-based multi-region network Feb. 2005 describe the termination of all their TE Link(s). This case is described in [GMPLS-ROUTING]. Moreover, in MRN, some network elements may be multiple switching capable nodes. Two types of multi-switching capable nodes are defined: simplex and hybrid nodes. - A simplex node, has a single switching capability per interface, but can comprise of interfaces with distinct switching capabilities (example: an LSR with PSC only interfaces and TDM only interfaces). - A hybrid node, on the other hand, has interfaces with multiple switching capabilities, and interfaces of the same hybrid node may have different multiple switching capabilities (ex. LSR with PSC + TDM interfaces). It may also have interfaces of a single switching capability, in addition to its interfaces supporting multiple switching capabilities. Simplex and Hybrid nodes can also be categorized according to the way they advertise these multiple switching capabilities. - A simplex node advertises several TE Links each with a single SC value as part of its ISCD sub-TLVs. - An hybrid node advertises at least one TE Link containing multiple ISCDs with different SC values (at least one per supported SC value per interface). Note: These cases are only partially described in [GMPLS- ROUTING]. 4.3.1 MRN with Simplex nodes In this case, the MRN network consists of at least one simplex node and include a set of single switching capable nodes (i.e. all TE links terminating on such nodes have the same switching capability). For example, the node TL2 in Figure 2 is a simplex node, which has links associated with TDM and links associated with LSC. At the region boundary, the interface switching capabilities of the ends of the link are different. When an LSP crosses the boundary from the upper to the lower regions, it is nested in a lower- region FA. ......................................................... : ........................................... : : : ............................. : : : : : ............... : : : : PSC : TDM : LSC : FSC : : : : : +--+ : +--+ : +--+ : +--+ +--+ : +--+ : +--+ : +--+ : : |P1|---|T1|---|L1|---|F1|---|F3|---|L3|---|T3|---|P3| : : +--+ : +--+ : +--+ : +--+ +--+ : +--+ : +--+ : +--+ : : | : | : | : | | : | : | : | : : | : | : | : | | : | : | : | : : +--+ : +---------+ : +--+ +--+ : +--+ : +--+ : +--+ : : |P2|---| TL2 |---|F2|---|F4|---|L4|---|T4|---|P4| : : +--+ : +---------+ : +--+ +--+ : +--+ : +--+ : +--+ : : : : .............. : : : : : ............................. : : : ........................................... : ......................................................... Figure 2: Simplex node MRN model. 4.3.2 MRN with hybrid nodes In this case, the MRN network consists of at least one hybrid node and include a set of single switching capable nodes (i.e. Shiomoto et al Expires August 2005 5 Requirements for GMPLS-based multi-region network Feb. 2005 all TE links terminating on such nodes have the same switching capability). Figure 3a shows an example of a hybrid node. The hybrid node has two switching elements, which have, for instance, interface switching capabilities PSC and TDM. It has two external interfaces (Link1 and Link2), which are directly connected to the switching element of TDM. The two switching elements are interconnected via an internal interface, which is not disclosed outside the network element. The internal interface is used to facilitate the "adaptation" between different switching capabilities: PSC and TDM. By cross-connecting port #a and port #b in the TDM switching element, if no reconfiguration of the ISCD sub-TLVs for Link 2 is performed, Link 2 is still advertised as being capable of TDM and PSC switching. Therefore, since there are no free resources for having a PSC FA Link terminating on this node, Link 2 should be advertised with a PSC ISCD sub-TLV with Max LSP bandwidth set to 0 for all priorities to described that only TDM resources are still available on this link. Network element ............................. : -------- : : | PSC | : : +--<->---| | : : | -------- : TDM : | ---------- : +PSC : +--<->--|#a TDM | : Link1 ------------<->--|#b | : Link2 ------------<->--|#c | : : ---------- : :............................ Figure 3a. Hybrid node. Figure 3b illustrates that existing GMPLS Routing is not sufficient and need to be extended to consider (internal) adaptation capabilities for hybrid nodes. Network element ............................. : -------- : : | PSC | : : | | : : --|#b1 | : : | | #d | : : | -------- : : | | : : | ---------- : : /| | | #c | : : | |-- | | : Link1 ========| | | TDM | : : | |----|#b2 | : : \| ---------- : :............................ Figure 3b. Hybrid node. Let's assume that all interfaces are STM64 (with VC4-16c capable as Max LSP bandwidth). So, initially, TE Link 1 composed is advertised with two ISCD sub-TLVs: - ISCD #1 sub-TLV: TDM with Max LSP bandwidth = STM16 (i.e. VC4- 16c capable as Max LSP bandwidth) and Unreserved bandwidth = STM64 - ISCD #2 sub-TLV: PSC with Max LSP bandwidth = 2.5 Gb (i.e. Shiomoto et al Expires August 2005 6 Requirements for GMPLS-based multi-region network Feb. 2005 and Unreserved bandwidth = 10 Gb After, setting up several PSC LSPs via port #d, there is only 155 Mb capacity still available between on port #d, however a 622 Mb capacity remains on port b1 and VC4-5c capacity in port b2. TE Link 1 is now advertised with the following ISCD sub- TLVs: - ISCD #1 sub-TLV: TDM with Max LSP bandwidth = VC4-4c, the Unreserved bandwidth reflects the VC4-5c capacity still available for port b2 - ISCD #2 sub-TLV: PSC with Max LSP bandwidth = 622 Mb, the Unreserved bandwidth reflects the capacity still available for the whole link i.e. 622 (port #b1) + 155 (port #d) Mb When computing the path for a new PSC LSP of 622 Mbps, one cannot know, based on existing GMPLS routing advertisements (i.e. two ISCD sub-TLVs), that it cannot setup a PSC-LSP that would be nested into a new VC4-4c TDM FA-LSP as there is only 155M still available for TDM-PSC adaptation. Thus, in that case additional routing information is required to advertise the available TDM- PSC internal adaptation resources (i.e. 155 Mb here). 4.3.3 Vertical and Horizontal interaction and integration Vertical interaction is defined as the collaborative mechanisms within a network element that is capable of supporting more than one switching capability (for example, PSC + LSC capable nodes). Integration of these interactions as part of the control plane is referred to as vertical integration. The latter refers thus to the collaborative mechanisms within a single control plane instance driving multiple switching capabilities (i.e. multiple LSP regions). Such a concept is useful in order to construct a framework that facilitates efficient network resource usage and rapid service provisioning in carrier's networks that are based on multiple switching technologies. Horizontal interaction is defined as the protocol exchange between network elements that support a single common switching technology (i.e. switching capability). For instance, the control plane interactions between two LSC network elements is an example of horizontal interaction. GMPLS protocol operations handle horizontal interactions within the same routing area. For the case where the interaction takes place across a domain boundary, such as between two routing areas that support the same switching technology, is currently being evaluated as part of the inter-domain work [Inter-domain] and referred to as horizontal integration. The latter refers thus to the collaborative mechanisms between network partitions and/or administrative boundaries such as routing areas or autonomous systems. This distinction gets blurred when administrative domains match LSP region boundaries. For example, the collaborative mechanisms in place between two lambda switching capable areas relate to horizontal integration. On the other hand, the collaborative mechanisms in place in a IP/MPLS over a TDM switching capable network could either be associated to horizontal integration (if each network is associated to an separate area) or to vertical integration (if both capabilities are located within the same area and driven by the same control plane instance). Networks where multiple switching capability (as defined in [RFC3945]) exist and are controllable through vertical interaction are termed "multi-region" networks. This document focuses on multi-region networks as a way to realize effective vertical integration. Shiomoto et al Expires August 2005 7 Requirements for GMPLS-based multi-region network Feb. 2005 4.3. Integrated Traffic Engineering (TE) and Resource Control In GMPLS-based multi-region networks, TE Links are consolidated into a single Traffic Engineering Database (TED). Since this TED contains the information relative to all the different regions existing in the network, a path across multiple regions can be computed using this TED. Thus optimization of network resources across the multiple regions can be achieved. The multi-region concept allows for the operation of one network switching type over another switching type (for example, the use of a PSC Forwarding Adjacency over an LSC network), the multi- region concept offers a greater degree of control and interworking including (but not limited too): - dynamic establishment of FA-LSPs - provisioning of end-to-end LSPs with dynamic FA-LSP triggering Note that MRN including multi-switching capable nodes, as the explicit route for establishing the end-to-end LSP can includes nodes belonging to multiple region (e.g. strict route), a mechanism to control the dynamic creation of FA-LSPs between each pair of node is required. There is a full spectrum of control as to how FA-LSPs can be established dynamically. It can be subject to the control of a policy, which may be set by a management component, and may require that the management plane is consulted at the time of FA establishment. Or FA-LSPs can be established at the request of the control plane without any management control. 4.4. Triggered signaling When a LSP crosses the boundary from an upper to a lower region, it may be nested in or stitched to a lower-region LSP. If such an LSP does not exist, the LSP may be established dynamically. Such a mechanism is referred to as "triggered signaling". 4.5. Forwarding adjacency (FA) Once an LSP across a lower region is created, it can be advertised as a TE-link called a Forwarding Adjacency (FA) TE link, allowing other nodes to use the LSP as a TE links for their path computation [HIER]. The FA is a useful and powerful tool for improving the scalability of GMPLS Traffic Engineering (TE) capable networks. The aggregation of TE Label Switched Paths (TE-LSPs) enables the creation of a vertical (nested) TE-LSP Hierarchy. A set of FAs across or within a lower region can be used by a higher region as part of the path computation process, and higher region LSPs may be carried across the FAs (just as they are carried across any TE link). This process requires either the nesting of LSPs through a hierarchical process [HIER]. In the MRN, since more than one higher region paths computation and modification can occur, FAs in the various regions are treated in a simple and efficient way. A MRN traffic engineering database (TED) is a set of FA information from multiple different regions. An FA's region is identified by the interface switching capability attached to the link state advertisement associated with the FA [GMPLS-ROUTING]. 4.6. Virtual network topology (VNT) A set of lower-region FAs provides a set of information for efficient path handling in the upper-region of the MRN, or provides a virtual network topology to the upper-region. For Shiomoto et al Expires August 2005 8 Requirements for GMPLS-based multi-region network Feb. 2005 instance, a set of FAs, each of which is instantiated by an LSC LSP, provides a virtual network topology to the PSC region, assuming that the PSC region is connected to the LSC region. The virtual network topology is configured by setting up or tearing down the LSC LSPs. By using GMPLS signaling and routing protocols, the virtual network topology can be easily adapted to traffic demands. By reconfiguring the virtual network topology according to traffic demand between source and destination node pairs, network performance factors, such as maximum link utilization and residual capacity of the network, can be optimized [MAMLTE]. Reconfiguration is performed by computing the new VNT from the traffic demand matrix and optionally from the current VNT. Exact details are outside the scope of this document. However, this method MAY be tailored according to the service provider's policy regarding network performance and quality of service (delay, loss/disruption, utilization, residual capacity, reliability). 5. Requirements 5.1. Requirements for multi-region TE 5.1.1 Scalability The MRN relies on a unified traffic engineering and routing model. The Traffic Engineering Database in each LSR will be populated with TE-links from all regions. This may lead to a huge amount of information that has to be flooded and stored within the network. Furthermore, path computation delays, which may be of huge importance during restoration, will depend on the size of the TE Database. Thus MRN routing mechanisms MUST be designed to scale well with an increase of any of the following: - Number of nodes - Number of TE-links (including FA-LSP) - Number of LSPs - Number of regions 5.1.2 FA link resource utilization It MUST be possible to utilize network resources efficiently. Particularly, resource usage in each region SHOULD be optimized as a whole (i.e. across all regions), in a coordinated manner. The number of lower-region FA-LSPs carrying upper-region LSPs SHOULD be minimized. Redundant lower-region FA-LSPs SHOULD be avoided (except for protection purpose). 5.1.2.1 FA release and setup Statistical multiplexing can only be employed in PSC and L2SC regions. PSC or L2SC (FA-)LSP may or may not consume the maximum reservable bandwidth of the FA-LSP. On the other hand, a TDM, or LSC (FA-)LSP always consumes a fixed amount of bandwidth of the FA-LSP as long as it exists (and is fully instantiated) because statistical multiplexing is not available. If there is low traffic demand, some FA-LSPs, which do not carry any LSP may be released so that resources are released. Note that if a small fraction of the available bandwidth is still under usage the nested LSPs can also be re-routed (make before break, before releasing the nesting FA-LSP. Alternatively, the FA-LSPs may be retained for future usage. Release or retention of underutilized FA-LSPs is a policy decision. Shiomoto et al Expires August 2005 9 Requirements for GMPLS-based multi-region network Feb. 2005 As part of the re-optimization process, the MRN solution MUST allow rerouting of FA-LSPs while keeping interface identifiers of FA links unchanged. Additional FAs MAY also be created based on policy, which might consider residual resources and the change of traffic demand across the region. By creating the new FAs, the network performance such as maximum residual capacity may be improved. As the number of FAs grows, the residual resource may decrease. In this case, re-optimization of FAs MAY be invoked according the policy. 5.1.2.2 Virtual FAs If FAs are used to enable connectivity over part or all of the lower-region, it may be considered disadvantageous to fully instantiate (i.e. pre-provision) the FA-LSPs since this may reserve bandwidth within the lower-region network that could be used for other LSPs in the absence of the upper-region traffic. However, in order that the upper-region can route traffic across the lower-region, the FA links MAY (this is not a MUST requirement as you can route an upper region LSP into a lower region based on lower region TE-links, even if there is no FA) still be advertised into the lower-region as TE links. Such FA links that represent the possibility of an FA-LSP are termed "virtual FAs". If an upper-region LSP that makes use of a virtual FA is set up, the underlying FA-LSP MUST be immediately signaled if it has not already been signaled. If virtual FAs are used in place of FAs, the TE links across the lower-region can remain stable using pre-computed paths while wastage of bandwidth within the lower-region, and unnecessary reservation of adaptation ports at the border nodes is avoided. The set of the virtual FAs defines the virtual FA topology across the lower region. The solution is expected to deliver the following mechanism in terms of the build-up of virtual FA topology operations taking into account the (forecast) traffic demand and available resource in the lower-region. The virtual FA topology MAY be modified dynamically (by adding or removing virtual FAs) according to the change of the (forecast) traffic demand and the available resource in the lower-region. The virtual FA topology can be changed by setting up and/or tearing down virtual FA-LSPs as well as by changes to real links and to real FAs. The maximum number of FAs that can be soft provisioned on a given resources SHOULD be well-engineered. How to design the virtual FA topology and its changes is out of scope of this document. 5.1.3 FA LSP Attribute inheritance FA TE-Link parameters SHOULD be inherited from FA-LSP parameters. These include: - Interface Switching Capability - TE metric - Maximum LSP bandwidth per priority level - Unreserved bandwidth for all priority levels - Maximum Reservable bandwidth - Protection attribute - Minimum LSP bandwidth (depending on the Switching Shiomoto et al Expires August 2005 10 Requirements for GMPLS-based multi-region network Feb. 2005 Capability) Inheritance rules MUST be applied based on specific policies. Particular attention should be given to the inheritance of TE metric and protection attributes. 5.1.4 Verify the FA before it enters service When the FA is created, it SHOULD be verified before it enters the in-service state. Data-plane connectivity, performance SHOULD be examined. 5.1.5 Disruption minimization When reconfiguring the virtual network topology according to the traffic demand change, the upper-region LSP may be disrupted. Such disruption MUST be minimized. When residual resource decreases to a certain level, some FAs may be released according to policies. Ideally, only FAs that are not carrying LSPs would be released, but in some cases it may be necessary to release FAs that are carrying traffic. 5.1.6 Path computation re-optimization stability When the virtual network topology is reconfigured, the path computation over the virtual network topology may be affected (re-optimized). The re-optimization of the path computation should be carefully controlled when the virtual network topology is reconfigured. The path computation is dependent on the network topology and associated link state. The path computation stability of upper region may be impaired if the Virtual Network Topology frequently changes and/or if the status and TE parameters (TE metric for instance) of links in the Virtual Network Topology changes frequently. In this context, robustness of the Virtual Network Topology is defined as the capability to smooth changes that may occur and avoid their subsequent propagation. Changes of the Virtual Network Topology may be caused by the creation and/or deletion of several LSPs. Creation and deletion of LSPs may be triggered by adjacent regions or through operational actions to meet change of traffic demand. Routing robustness should be traded with adaptability with respect to the change of incoming traffic requests. A full mesh of LSPs may be created between every pair of border nodes of the PSC region. The merit of a full mesh of PSC FAs is that it provides stability to the PSC-level routing. That is, the forwarding table of an PSC-LSR is not impacted by re-routing changes within the lower-region (e.g., TDM). Further, there is always full PSC reachability and immediate access to bandwidth to support PSC LSPs. But it also has significant drawbacks, since it requires the maintenance of n^2 RSVP-TE sessions, which may be quite CPU and memory consuming (scalability impact). 5.1.7 Computing paths with and without nested signaling Path computation may take into account LSP region boundaries when computing a path for an LSP. For example, path computation may restrict the path taken by an LSP to only the links whose interface switching capability is PSC-1. Shiomoto et al Expires August 2005 11 Requirements for GMPLS-based multi-region network Feb. 2005 Interface switching capability is used as a constraint in computing the path. A TDM-LSP is routed over the topology composed of TE links, both of whose ends has TDM switching capability. In Figure 4, a TDM-LSP is routed from LSR-P1, through TDM_SW-T1 and TDM_SW-T2, to LSR-P2. The path for the TDM-LSP is composed of links, both of whose ends has TDM switching capability. Once the TDM LSP is set up, it is advertised as an FA-LSP, both ends of which are PSC. In calculating the path for the PSC-LSP, the TE database is filtered to include the link, both ends of which include only PSC. In this way hierarchical routing of the PSC- LSP and TDM-LSP is done by using a TE database filtered with respect to switching capability. There may be a case, in which we can set up the LSP if we build new lower-region LSPs along the computed path. Suppose that we set up the TDM-LSP between P1 and P2 in Figure 5. The TDM-LSP is routed over the path T1-L1-L2-T2. At this time, there is no direct link between T1 and T2. Then, the LSC-LSP is set up between T1 and T2. The LSC-LSP setup request (between T1 and T2) is triggered by the TDM-LSP setup request (between P1 and P2). If triggered signaling is allowed, the path computation mechanism may produce a route containing multiple regions. .................................................. : .................................. : : : ................. : : : : : : : : : PSC : TDM : LSC : : : : +--+ : +--+ : +--+ +--+ : +--+ : +--+ : : |P1|-----|T1|-----|L1|---|L2|-----|T2|----|P2| : : +--+ : +--+ : +--+ +--+ : +--+ : +--+ : : : | ................. | : : : : | | : : : : ------------------------ : : : .................................. : .................................................. Figure 4 and 5. Path computation in MRN. 5.1.8 Handling single-switching and multi-switching TE links The MRN can consist of single-switching capable and multi switching capable TE links. The path computation mechanism in the MRN SHOULD be able to compute the paths consisting of both types of TE-links. Recall the simplex node model shown in Figure 2. The switching capability of both ends of a TE-link may or may not be the same. For a TE-link between an LSR and a TDM switch, the switching capability of the end-point on the LSR-side is PSC while the one on the TDM switch-side is TDM. For a TE-link between two TDM switches, the switching capability of the both end-points is TDM. The links of the hybrid node shown in Figure 3 are advertised as TE-links with multiple interface switching capabilities: PSC and TDM. The hybrid node is used as a transit node for a TDM-region. At the same time, the hybrid node is used as an ingress, egress, or transit node for the PSC-region. 5.1.9 Advertisement of the available adaptation resource A multi-switching capable node is required to hold and advertise resource information on its internal links. Shiomoto et al Expires August 2005 12 Requirements for GMPLS-based multi-region network Feb. 2005 For example, if the hybrid node shown in Figure 3a is used as an ingress or egress node, once the cross-connection is made between port #a and #b in the TDM switching element, a new FA link is advertised as with a single switching capability: PSC. After that, there is no available internal link to connect port #b to the PSC. Therefore, a mechanism is required such that Link 2 ISCD sub-TLVs are advertised with Max LSP bandwidth values reflecting that only TDM resources are still available on this link. However, as shown in Figure 3b, the above mechanism is not realizable anymore when a given switching capability is accessed directly from the incoming link and from another switching capability hosted by the same node. Therefore, within multi- region networks, the advertisement of the so-called adaptation capability to terminate LSPs is required, as it provides critical information when performing multi-region path computation. 6. Security Considerations The current version of .his document does not introduce any new security considerations as it only lists a set of requirements. In the future versions, new security requirements may be added. 7. References 7.1. Normative Reference [MPLSGMPLS] D.Brungard, J.L.Le Roux, E.Oki, D. Papadimitriou, D.Shimazaki, K.Shiomoto, "Migrating from IP/MPLS to GMPLS networks," draft-oki-ccamp-gmpls-ip- interworking-03.txt (work in progress), July 2004. [GMPLS-ROUTING]K.Kompella and Y.Rekhter, "Routing Extensions in Support of Generalized Multi-Protocol Label Switching," draft-ietf-ccamp-gmpls-routing-09.txt, October 2003 (work in progress). [Inter-domain] A.Farrel, J-P. Vasseur, and A.Ayyangar, "A framework for inter-domain MPLS traffic engineering," draft-ietf-ccamp-inter-domain- framework-00.txt, work in prgoress, July 2004. [HIER] K.Kompella and Y.Rekhter, "LSP hierarchy with generalized MPLS TE," draft-ietf-mpls-lsp- hierarchy-08.txt, work in progress, Sept. 2002. [LMP] J. Lang, "Link management protocol (LMP)," draft- ietf-ccamp-lmp-10.txt (work in progress), October 2003. [RFC3945] E.Mannie (Ed.), "Generalized Multi-Protocol Label ` Switching (GMPLS) Architecture", RFC 3945, October 2004. 7.2. Informative References [MAMLTE] K. Shiomoto et al., "Multi-area multi-layer traffic engineering using hierarchical LSPs in GMPLS networks", draft- shiomoto-multiarea-te-01.txt (work in progress). 8. Author's Addresses Kohei Shiomoto NTT Network Service Systems Laboratories Shiomoto et al Expires August 2005 13 Requirements for GMPLS-based multi-region network Feb. 2005 3-9-11 Midori-cho, Musashino-shi, Tokyo 180-8585, Japan Email: shiomoto.kohei@lab.ntt.co.jp Dimitri Papadimitriou Alcatel Francis Wellensplein 1, B-2018 Antwerpen, Belgium Phone : +32 3 240 8491 Email: dimitri.papadimitriou@alcatel.be Jean-Louis Le Roux France Telecom R&D, Av Pierre Marzin, 22300 Lannion, France Email: jeanlouis.leroux@francetelecom.com Martin Vigoureux Alcatel Route de Nozay, 91461 Marcoussis cedex, France Phone: +33 (0)1 69 63 18 52 Email: martin.vigoureux@alcatel.fr Deborah Brungard AT&T Rm. D1-3C22 - 200 S. Laurel Ave., Middletown, NJ 07748, USA Phone: +1 732 420 1573 Email: dbrungard@att.com Contributors Eiji Oki (NTT Network Service Systems Laboratories) 3-9-11 Midori-cho, Musashino-shi, Tokyo 180-8585, Japan Phone: +81 422 59 3441 Email: oki.eiji@lab.ntt.co.jp Ichiro Inoue (NTT Network Service Systems Laboratories) 3-9-11 Midori-cho, Musashino-shi, Tokyo 180-8585, Japan Phone: +81 422 59 3441 Email: ichiro.inoue@lab.ntt.co.jp Emmanuel Dotaro (Alcatel) Route de Nozay, 91461 Marcoussis cedex, France Phone : +33 1 6963 4723 Email: emmanuel.dotaro@alcatel.fr 9. Intellectual Property Considerations The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79. By submitting this Internet-Draft, each author represents that any applicable patent or other IPR claims of which he or she is aware have been or will be disclosed, and any of which he or she becomes aware will be disclosed, in accordance with Section 6 of RFC 3668. Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this Shiomoto et al Expires August 2005 14 Requirements for GMPLS-based multi-region network Feb. 2005 specification can be obtained from the IETF on-line IPR repository at http://www.ietf.org/ipr. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf-ipr@ietf.org. The IETF has been notified by Tellabs Operations, Inc. of intellectual property rights claimed in regard to some or all of the specification contained in this document. For more information, see http://www.ietf.org/ietf/IPR/tellabs-ipr-draft- shiomoto-ccamp-gmpls-mrn-reqs.txt 10. Full Copyright Statement Copyright (C) The Internet Society (2005). This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights. This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Shiomoto et al Expires August 2005 15