| PCE Working Group | R. Casellas, Ed. |
| Internet-Draft | CTTC |
| Intended status: Informational | C. Margaria |
| Expires: July 14, 2014 | Coriant |
| A. Farrel | |
| Old Dog Consulting | |
| O. Gonzalez de Dios | |
| Telefónica I+D | |
| D. Dhody | |
| X. Zhang | |
| Huawei Technologies | |
| January 10, 2014 |
Current issues with existing RBNF notation for PCEP messages and extensions
draft-cmfg-pce-pcep-grammar-02
The PCEP protocol has been defined in [RFC5440] and later extended in several RFCs. This document aims at documenting inconsistencies when implementing a set of extensions and at providing a reference, complete and formal RBNF grammar for PCEP messages, including object ordering and precedence rules.
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 July 14, 2014.
Copyright (c) 2014 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.
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 [RFC2119].
The RBNF notation, defined in [RFC5511], is used to specify the message format for the Path Computation Element Communication Protocol (PCEP). The core of PCEP has been defined in [RFC5440] and later extended in [RFC5441], to support the Backward Recursive Path Computation (BRPC) procedure; in [RFC5455], adding a CLASSTYPE object to support Diffserv-aware Traffic Engineering (DS-TE); in [RFC5520], for topology confidentiality by means of Path keys; in [RFC5521], in support of exclusions; in [RFC5541] to convey specific objective functions; in [RFC5557], for Global Concurrent Optimization, in [RFC5886], for monitoring and in [RFC6006] for point-to-multipoint (P2MP) computation.
Most PCEP RFCs describe specific protocol extensions and, as such, they focus on their constructs extending some base RFCs. Although it is not the intention of each individual draft or RFC to provide the latest and most complete/full definition of the protocol messages, in practice combining all the extensions as defined in the respective RFCs is complex.
Message rules are sometimes provided within the text, resulting in ambiguity. Moreover, the fact that extensions may be defined in parallel may be a problem. The canonical example is the case where RFC X defines construct p ::= A and subsequent RFC Y extends RFC X stating that object C MUST follow object A and RFC Z also extends RFC X stating that object D MUST follow object A.
The use of RBNF [RFC5511] states that the ordering of objects and constructs in an assignment is explicit, and protocol specifications MAY opt to state that ordering is only RECOMMENDED (the elements of a list of objects and constructs MAY be received in any order).
The core PCEP document [RFC5440] states in Section 6 that an implementation MUST form the PCEP messages using the object ordering specified in [RFC5440].
[RFC5886] equally states that "An implementation MUST form the PCEP messages using the object ordering specified in this document."
[RFC5521] only states that "the XRO is OPTIONAL and MAY be carried within Path Computation Request (PCReq) and Path Computation Reply (PCRep) messages." and no ordering is provided. It does not mention SVEC objects or rules.
[RFC5541] specifies that "the OF object MAY be carried within a PCReq message. If an objective function is to be applied to a set of synchronized path computation requests, the OF object MUST be carried just after the corresponding SVEC (Synchronization VECtor) object and MUST NOT be repeated for each elementary request. Similarly, if a metric is to be applied to a set of synchronized requests, the METRIC object MUST follow the SVEC object and MUST NOT be repeated for each elementary request. (...) An OF object specifying an objective function that applies to an individual path computation request (non-synchronized case) MUST follow the RP object for which it applies". It should be understood that this last sentence must be relaxed or is in contradiction with the ENDPOINTS object.
RFCs that extend the core PCEP protocol are not consistent with the object ordering. For example, [RFC5520] defines:
<segment-computation> ::=
<END-POINTS>
[<LSPA>]
[<BANDWIDTH>]
[<BANDWIDTH>]
[<metric-list>]
(snip)
<request>::= <RP>
<END-POINTS>
[<LSPA>]
[<BANDWIDTH>]
[<metric-list>]
[<RRO>[<BANDWIDTH>]]
[<IRO>]
[<LOAD-BALANCING>]
and states that "the format of the message for use in normal path computation is unmodified". However, [RFC5520] was not updated to reflect that the the BANDWIDTH object used for reoptimization was moved to appear after the RRO for which it applies, as given in [RFC5440] (updated in Errata ID: 3582):
[RFC5541] in section 3.2 is not consistent with the ordering of OF and metric-list:
<svec-list> ::= <SVEC>
[<OF>]
[<metric-list>]
<request> ::= <RP>
(snip)
[<metric-list>]
[<OF>]
<attribute-list> ::= [<OF>]
[<LSPA>]
[<BANDWIDTH>]
[<metric-list>]
In view of the above considerations, this document aims at providing an object ordering for PCEP messages so implementations can interoperate. Implementations conforming to this document MUST use the object ordering specified here.
PCEP RFCs may use inconsistent or ambiguous naming. For example [RFC5440] defines the Open message as having a common header and an OPEN object, and later uses Open to refer to the object that may appear in a PCErr message.
<Open Message> ::= <Common Header>
<OPEN>
<PCErr Message> ::= <Common Header>
(<error-obj-list> [<Open>]) | <error>
[<error-list>]
It is common that a sequence or repetition of an object OBJ is noted as obj-list. It may happen that in extensions to core documents, the naming is kept although it no longer applies to such a sequence. For example, [RFC5886] states:
<svec-list> ::= <SVEC>
[<OF>]
[<svec-list>]
and later
<svec-list> ::= <SVEC>
[<svec-list>]
The current RBNF notation does not capture the semantics/intent of the messages; notably, when two options are mutually exclusive and at least one is mandatory. In most cases, this is noted as both options being optional. For example [RFC5440] states:
<response>::=<RP>
[<NO-PATH>]
[<attribute-list>]
[<path-list>]
with this example, a message that contains a response of the form <RP><NO-PATH><ERO><..> (that is, a NO-PATH object followed by a path) is correct and successfully parsed. Likewise, a response with just an RP object is valid. Although the actual text within the RFC may state the intention and disambiguate the grammar, having a RBNF notation that better captures semantics, message structure and original intent, enables the development of automated parsers that closely map the specification.
Similarly, if the intent is to specific a rule such as metric-pce which includes a PCE-ID object followed by a PROC-TIME object and/or an OVERLOAD object, the syntax:
<metric-pce> ::= <PCE-ID> [<PROC-TIME>] [<OVERLOAD>]
allows, amongst other combinations, that neither PROC-TIME nor OVERLOAD appears, which is not the intended behavior (there should be at least one metric). The alternative
<metric-pce> ::= <PCE-ID> <metric-argument-list> <metric-argument-list> ::= <metric-argument> [<metric-argument-list>] <metric-argument> ::= <PROC-TIME> | <OVERLOAD>
or equivalently
<metric-pce> ::= <PCE-ID> (<metric-argument>...) <metric-argument> ::= <PROC-TIME> | <OVERLOAD>
does not reflect that each metric-argument should appear at most once. This can be addressed verbosely:
<metric-pce> ::= <PCE-ID>
( <PROC-TIME> | <OVERLOAD> | <PROC-TIME><OVERLOAD> )
<metric-pce> ::= <PCE-ID>
( <PROC-TIME>[<OVERLOAD>] | [<PROC-TIME>]<OVERLOAD> )
Here the semantic is that we require any object of the set {PROC-TIME, OVERLOAD} to be present, and there should be at least one. Note that currently there are only a few cases where the "non-empty set" case arises.
- non-empty set, repetition not allowed
<set> ::= { <a> | <b> | <c> }
- non-empty set, repetition allowed
<set> ::= { <a> <b> <c> }
-- also can be expressed using the previous
definition with
<set> ::= { <a>... | <b>... | <c>... }
Note that the other options can already be handled
- non-repetition set allowed to be empty
<set> ::= [<a>] [<b>] [<c>]
- repetition set allowed to be empty
<set> ::= [<a>] [<b>] [<c>] [<set>]
The notation with "{" would be convenient to express implicit
ordering (<a><a><b> ok but <a><b><a> not)].
[Editor note/AF To make a normative or machine-readable definition, new notation could be defined:
A more condensed notation extension to the RBNF notation could also use a "sequential or" notation:
<a> || <b> is defined as <a> | <b> | <a><b>
<a> || <b> || <c> is defined as (assoc.)
(<a> | <b> | <a> <b>) | <c> | (<a> | <b> | <a> <b> ) <c> =
(<a> | <b> | <a> <b>) | <c> | (<a><c> | <b><c> | <a><b><c>) =
<a> | <b> | <c> | <a> <b> | <a> <c> | <b> <c> | <a> <b> <c>
The use of sequential-or notation allows writing:
<metric-pce> ::= <PCE-ID> ( <PROC-TIME> || <OVERLOAD> )
The goal of this document is then, first, to provide an (almost) formal (reasonably) complete definition of PCEP messages, checking the overall protocol and extensions consistency, defining an object ordering; and to set the basis for implementation agreements that aim at integrating published PCEP extensions. It is also a goal to provide alternative (although compatible) RBNF notations to be expressive enough to avoid invalid cases.
This document does not modify the content of defined PCEP objects and TLVs.
This document is not normative, the normative definition is included in the existing specs. This does not preclude integration with a future revision of such documents.
This section provides the proposed RBNF notation for the PCEP messages. Specific constructs or grammar rules that appear in several messages or deserve special considerations are described first.
<of-list> ::= <OF> [<of-list>]
<metric-list> ::= <METRIC> [<metric-list>]
<vendor-info-list> ::= <VENDOR-INFORMATION> [<vendor-info-list>]
<pce-id-list> ::= <PCE-ID> [<pce-id-list>]
-- (note: named pce-list in original)
<svec-tuple> ::= <SVEC>
[<OF>]
[<metric-list>]
[<vendor-info-list>]
[<GC>]
[<XRO>]
<svec-tuple-list> ::= <svec-tuple> [<svec-tuple-list>]
Note that [I-D.ietf-pce-vendor-constraints] defines:
<svec-list> ::= <SVEC>
[<OF>]
[<GC>]
[<XRO>]
[<metric-list>]
[<vendor-info-list>]
[<svec-list>]
The construct is updated to reflect the new name and to have the same relative order in the attributes that constrain a inidivudal request
A metric-pce-id is a rule that associates a PCE identified by its PCE-ID to a list of metric arguments.
<metric-pce-id> ::= <PCE-ID>
(<PROC-TIME> [<OVERLOAD>] |
[<PROC-TIME>] <OVERLOAD> )
<metric-pce-id-list> ::= <metric-pce-id> [<metric-pce-id-list>]
See [RFC5886] for the definition of specific/general and in-band/out-of-band.
<monitoring> ::= <MONITORING> <PCC-ID-REQ>
<monitoring-request> ::= <monitoring> [<pce-id-list>]
<monitoring-response> ::= <monitoring>
(<specific-monitoring-metrics-list> |
<general-monitoring-metrics-list>)
<specific-monitoring-metrics-list> ::=
<specific-monitoring-metrics>
[<specific-monitoring-metrics-list>]
<general-monitoring-metrics-list> ::=
<general-monitoring-metrics>
[<general-monitoring-metrics-list>]
<specific-monitoring-metrics> ::=
<RP> <monitoring-metrics>
<general-monitoring-metrics> ::=
<monitoring-metrics>
<monitoring-metrics> ::=
<metric-pce-id-list>
<Open Message> ::= <Common Header>
<OPEN>
<KeepAlive Message>::= <Common Header>
Note that the actual parsing depends on the content (flags) of the Request Parameters (RP) object, notably expansion and P2MP. In some cases, this may be considered redundant, e.g. the presence of a PATH_KEY object and the corresponding flag.
<a with x=v> and <a with x!=v> then (<a with x=v> <b>) | (<a with x!=v> <c>)
[Editor's note: rom a notation perspective, we lack a way to express "if object a field x has value v then include object b, else include object c". A possible way would be to define new intermediate types :
The PCReq message contains a possibly monitored list of requests, some of which may be grouped by SVEC tuples.
<PCReq Message>::= <Common Header>
[<monitoring-request>]
[<svec-tuple-list>]
<request-list>
where:
<request-list> ::= <request> [<request-list>]
-- A request is either an expansion, a P2P request or a P2MP request
<request> ::= <expansion> |
<p2p_computation> |
<p2mp_computation>
<expansion> ::= <RP><PATH-KEY>
<p2p_computation> ::= <RP><ENDPOINTS>
[<LSP>][<gen-bw>][<p2p-attributes>...]
<p2mp_computation> ::= <RP><tree-list>
[<p2mp-attributes>...]
-- For a P2P computation
<p2p-attributes> ::= <attributes>|<rro-bw-pair>
<attributes> ::= <attribute> [<attributes>]
<attribute> ::=
<CLASSTYPE> |
<LSPA> |
<OF> |
<metric-list> |
<vendor-info-list> |
<IRO> |
<BNC> |
<XRO> |
<gen-load-balancing> |
<INTER-LAYER> |
<SWITCH-LAYER> |
<REQ-ADAP-CAP>
-- in RFC6006 there is a bw per tree,
-- it is intended to be an optimization for an RRO list
<rro-bw-pair> ::= <RRO> [<gen-bw>]
<rro-list-bw> ::= (<RRO>...)[<gen-bw>]
<tree> ::= <ENDPOINTS>(<rro-bw-pair>|<rro-list-bw>)
<gen-bw> ::= <BANDWIDTH>[<GENERALIZED-BANDWIDTH>...]
-- per RFC5440 section 7.7
<gen-load-balancing> ::= <LOAD-BALANCING> |
<GENERALIZED-LOAD-BALANCING>
-- For P2MP computations - note some atts (BNC) are only P2MP
<tree-list> ::= <tree> [<tree-list>]
<tree> ::= <ENDPOINTS> <rro_bw_pair>
<p2mp-attributes> ::= (<attribute> | <BNC>) [<p2mp-attributes>]
<PCRep Message> ::= <Common Header>
[<svec-tuple-list>]
<response-list>
-- Note: should clarify the use of SVEC tuple list
where
<response-list> ::= <response> [<response-list>]
-- An individual response may include monitoring info
<response> ::= <RP> [<monitoring>]
(<success> | <failure>) [<monitoring-metrics>]
-- Note: should clarify P2MP attributes
<success> ::= <path-list>
<failure> ::= <NO-PATH> [<attributes>]
<path-list> ::= <path>[<path-list>]
<path> ::= <ERO> <gen-bw> [<attributes>]
<PCMonReq Message> ::= <Common Header>
<monitoring-request>
[[<svec-tuple-list>] <request-list>]
<PCMonRep Message> ::= <Common Header>
<monitoring-response>
<PCNtf Message> ::= <Common Header> ( <solicited-notify> | <unsolicited-notify> ) where <solicited-notify> ::= <request-id-list> <notification-list> <unsolicited-notify> ::= <notification-list> <request-id-list> ::= <RP> [<request-id-list>] <notification-list> ::= <NOTIFICATION> [<notification-list>]
<PCErr Message> ::= <Common Header>
( <solicited-error> | <unsolicited-error> )
where
<solicited-error> ::= <request-id-list> <pcep-error-list>
<unsolicited-error> ::= <handshake-error> | <pcep-error-list>
<handshake-error> ::= <pcep-error-list> <OPEN>
<request-id-list> ::= <RP> [<request-id-list>]
<pcep-error-list> ::= <PCEP-ERROR> [<pcep-error-list>]
TBD see [I-D.ietf-pce-stateful-pce].
TBD see [I-D.ietf-pce-stateful-pce].
TBD
This work was supported in part by the PACE Support Action (http://ict-pace.net/) project under grant agreement number 619712.