Folks,

Please note that draft 12 is the same as draft 13 except for the end-of-line characters. So you should compare 13 to 11 (or an earlier version depending on when you last reviewed the document.)

NOTE: Version 13 has end-of-line characters (linefeeds) that Rfcdiff will handle. However, as uploaded, versions 9, 10, 11, and 12 have carriage returns as their end-of-line characters and don't work with Rfcdiff. I believe that if you substitute LF for all the CRs, Rfcdiff will work with them. I'm told that downloading with IE automatically does the end-of-line changes. And Word let me select the end-of-line terminator when I used it to open and then save the files.

In addition to some clarifications, addition of references, and correction of typos and minor errors, the following revisions were made between draft 11 and 12/13 to address feedback from Tim Polk and Sean Turner.

Karen

1. Introduction

From: Note: This document is based on the template specified in the
   Internet Engineering Task Force (IETF) standards document RFC 3647.
   A number of sections contained in the template were omitted from
   this policy because they did not apply to this PKI. However, we have
   retained the section numbering scheme employed in the RFC to
   facilitate comparison with the outline in Section 6 of the RFC.

To:   Note: This document is based on the template specified in the
   Internet Engineering Task Force (IETF) standards document RFC 3647
   [RFC3647].  In the interest of keeping the document as short as
   reasonable, a number of sections contained in the template are
   omitted from this policy because they did not apply to this PKI.
   However, we have retained the section numbering scheme employed in
   the RFC to facilitate comparison with the outline in Section 6 of
   the RFC. Each of these omitted sections should be read as "No
   stipulation" in CP/CPS parlance.


1.2. Overview -- replaced 2nd paragraph with 2 paragraphs

From: This PKI parallels the existing INR distribution hierarchy. These
   resources are distributed by the Internet Assigned Numbers Authority
   (IANA) to the Regional Internet Registries. In some regions,
   National Internet Registries (NIRs) form a tier of the hierarchy
   below the RIRs for internet number resource (INR) distribution. ISPs
   and network subscribers form additional tiers below registries.

To:   The structure of the RPKI is modeled on the existing organizational
   structure that is already responsible for IP address and AS number
   resource allocation. In this allocation hierarchy, IANA allocates
   resources to five Regional Internet Registries (RIRs), each of which
   manages address and AS number allocation within a defined
   geopolitical region. The RIRs in turn allocate resources to Internet
   Service Providers, to subscribers with so-called provider-
   independent ("portable") allocations, and in some regions, to
   National Internet Registries (NIRs). (The term LIR is used in some
   regions to refer to what other regions define as an ISP. Throughout
   the rest of this document we will use the term ISP to simplify
   references to these entities.)

   As noted in the RPKI architecture document [ARCH], although IANA
   and/or the five RIRs are obvious candidates to be default TAs for
   the RPKI, each RP ultimately chooses the set of trust anchors it
   will use for certificate validation. For example, an RP (e.g., an
   LIR/ISP) could create a trust anchor to which all address space
   and/or all AS numbers are assigned, and for which the RP knows the
   corresponding private key. The RP could then issue certificates
   under this trust anchor to whatever entities in the PKI it wishes,
   with the result that the certification paths terminating at this
   locally-installed trust anchor will satisfy the RFC 3779 validation
   requirements.


1.4.1. Certification authorities -- 2nd paragraph

From: Organizations that do not distribute INRs, but hold such resources
   also act as CAs when they create EE certificates.

To:   Organizations that do not distribute INRs, but hold such resources
   also act as CAs when they create EE certificates. Subscribers with
   PI space are the principle examples of this class of CAs.


1.6.4. CP approval procedures

From: IESG MUST approve a replacement BCP, following the procedures of the
   IETF Standards Process as defined in RFC 2026.

To:   The IESG MUST approve a replacement BCP that either updates or
   obsoletes this BCP, following the procedures of the IETF Standards
   Process as defined in RFC 2026 [RFC2026].


1.7 Definitions and acronyms

Added:
       LIR -  In some regions, the term Local Internet Registry (LIR) is
    used to refer to what is called an ISP in other regions.

       Provider-independent (PI) address space - PI address space is
    allocated directly by an Internet Registry to a subscriber. This space
    cannot be further allocated by the subscriber to other entities. The
    subscriber must contract with an ISP to obtain routing of the address
    block within the Internet. PI space enables the subscriber to change
    service providers without renumbering its networks and to use multiple
    ISPs in a multi-homed configuration.


2.2 Publication of certification information

Added:
    Each CA MUST publish its CPS.


3.1.2 Need for names to be meaningful

From: The Subject name in each certificate MUST be unique relative to all
   Subject names certified by the same Issuer, but the name does not
   need to be meaningful. There is no requirement, and no guarantee,
   that subject names are globally unique in this PKI. The certificates
   issued under this PKI are used for authorization in support of
   applications that make use of attestations of Internet resource
   holding, not for identification. The PKI binds a public key to sets
   of INRs. The name of the holder of the INR is not intended to be
   "meaningful".

To:   The Subject name in each certificate does not need to be meaningful.
   The certificates issued under this PKI are used for authorization in
   support of applications that make use of attestations of Internet
   resource holdings, not for identification. The PKI binds a public
   key to sets of INRs. The name of the holder of the INR is not
   intended to be "meaningful".


3.1.5. Uniqueness of names

From: Each CA certifies Subject names that MUST be unique among the
   certificates that it issues. Although it is desirable that these
   Subject names be unique throughout the PKI, to facilitate
   certificate path discovery, such uniqueness is neither mandated nor
   enforced through technical means.

To:   There is no guarantee that subject names are globally unique in this
   PKI. Each CA certifies Subject names that MUST be unique among the
   certificates that it issues. Although it is desirable that these
   Subject names be unique throughout the PKI, name uniqueness within
   the RPKI cannot be guaranteed.

   However, Subject names in certificates SHOULD be constructed in a
   way that minimizes the chances that two entities in the RPKI will be
   assigned the same name.


3.1.6. Recognition, authentication, and role of trademarks

From: Because the Subject names are not intended to be meaningful, there
   is no provision to recognize or authenticate trademarks, service
   marks, etc.

To:   Since Subject names are not intended to be meaningful, there is no
   provision to recognize or authenticate trademarks, service marks,
   etc. When generating subject names, CAs should use a mechanism that
   is unlikely to result in Subject names that include trademarks,
   service marks, etc. CAs MUST NOT knowingly issue a certificate with
   a Subject name that a court of competent jurisdiction has determined
   infringes on the trademark of another individual or organization.


4.2. Certificate application processing

From: CAs SHOULD make use of existing standards for certificate
   application processing.  Relevant standards include RFC 4210,
   Internet X.509 Public Key Infrastructure Certificate Management
   Protocol (CMP), RFC 2797, Certificate Management Messages over CMS,
   and RSA Labs standards PKCS #7 and PKCS #10. Each CA MUST define the
   certificate request/response standards that it employs, via its CPS.

To:   CAs SHOULD make use of existing standards for certificate
   application processing. Section 6 of the resource certificate
   profile [RFCyyyy] defines the standard certificate request formats
   that MUST be supported

   Each CA MUST define the certificate request/response standards that
   it employs, via its CPS.


4.5.2. Relying party public key and certificate usage -- last paragraph

From: If a relying party determines that use of the certificate is
   appropriate, the relying party must utilize appropriate software
   and/or hardware to perform digital signature verification as a
   condition of relying on the certificate. Moreover the relying party
   MUST validate the certificate in a manner consistent with the
   algorithms specified in RFC 5280 and RFC 3779.

To:   If a relying party determines that use of the certificate is
   appropriate, the relying party must utilize appropriate software
   and/or hardware to perform digital signature verification as a
   condition of relying on the certificate. Moreover the relying party
   MUST validate the certificate in a manner consistent with the RPKI
   certificate profile [RFCyyyy], which specifies the extended
   validation algorithm for RPKI certificates.


4.6 Certificate renewal

Added: This section describes the procedures for certificate renewal. Certificate renewal is the issuance of a new certificate to replace an old one prior to its expiration. Only the validity dates and the serial number are changed. The public key and all other information remain the same.


4.6.1. Circumstance for certificate renewal -- first paragraph.

Added: An overlap period of at least 24 hours is RECOMMENDED.


Sections 4.X.4, 4.X.5, 4.X.6, 4.X.7

Made consistent with each other as follows:

4.6.4, 4.7.4, 4.8.4 Notification of new certificate issuance to subscriber
        No additional stipulations beyond those of section 4.3.2.

4.6.5, 4.7.5, 4.8.5 Conduct constituting acceptance of a [renewal, re-keyed, modified] certificate
        No additional stipulations beyond those of section 4.4.1.

4.6.6, 4.7.6, 4.8.6 Publication of the [renewal, rekeyed, modified] certificate by the CA
        No additional stipulations beyond those of section 4.4.2.

4.6.7, 4.7.7, 4.8.7 Notification of certificate issuance by the CA to other entities
        No additional stipulations beyond those of section 4.3.3.


4.7. Certificate re-key

Added: This section describes the procedures for certificate re-key. Certificate re-key is the issuance of a new certificate to replace an old one because the key needs to be replaced. Unlike with certificate renewal, the public key is changed.


4.7.1. Circumstance for certificate re-key -- 3rd paragraph

From: Note that if a certificate is revoked to replace the RFC 3779
   extensions, the replacement certificate MUST incorporate the same
   public key rather than a new key.

To:    Note that if a certificate is revoked to replace the RFC 3779
   extensions, the replacement certificate MUST incorporate the same
   public key rather than a new key.  This applies to when one is
   adding INRs (revocation not required) and to when one is removing
   INRs (revocation required (see Section 4.8.1).


4.9.3. Procedure for revocation request

From: A subscriber MAY submit a request to the certificate issuer for a
   revocation. A certificate issuer MUST notify the subscriber when
   revoking a certificate, however this notification requirement MAY be
   satisfied by publication of a CRL by the issuer. If means other than
   CRL publication are employed, the CPS for a CA MUST indicate by what
   means the CA will inform a subscriber of certificate revocation,

To:   A subscriber MAY submit a request to the certificate issuer for a
   revocation. This request MUST identify the certificate to be revoked
   and MUST be authenticated. The procedures for making the request
   MUST be described in the CPS for each CA. The RPKI provisioning
   document [PROV] describes a protocol that MAY be used to make
   revocation requests.

   A certificate issuer MUST notify the subscriber when revoking a
   certificate. The notification requirement is satisfied by CRL
   publication. The CPS for a CA MUST indicate the means by which the
   CA will inform a subscriber of certificate revocation.


4.10. Certificate status services

From: This PKI does not make provision for use of OCSP or SCVP, because it
   is anticipated that the primary relying parties (ISPs) will acquire
   and validate certificates for all participating resource holders on
   a daily basis. These protocols are not designed for such large-
   scale, bulk certificate status checking. Instead, retrieval of all
   changed or new certificates and CRLs on a daily basis is the
   anticipated mode of certificate status verification.

To:   This PKI does not make provision for use of OCSP or SCVP, because it
   is anticipated that the primary RPs (ISPs) will acquire and validate
   certificates for all participating resource holders. These protocols
   are not designed for such large-scale, bulk certificate status
   checking. RPs MUST check for new CRLs at least daily. It is
   RECOMMENDED that RPs perform this check several times per day, but
   no more than 8-12 times per day (to avoid excessive repository
   accesses).


5.1. Physical controls

Added:  Possible topics include:

     .  Site location and construction, such as the construction
        requirements for high-security zones and the use of locked
        rooms, cages, safes, and cabinets

     .  Physical access, i.e., mechanisms to control access from one
        area of the facility to another or access into high-security
        zones, such as locating CA operations in a secure computer room
        monitored by guards or security alarms and requiring movement
        from zone to zone to be accomplished using a token, biometric
        readers, and/or access control lists

     .  Power and air conditioning

     .  Water exposures

     .  Fire prevention and protection

     .  Media storage, for example, requiring the storage of backup
        media in a separate location that is physically secure and
        protected from fire and water damage

     .  Waste disposal

     .  Off-site backup


5.2. Procedural controls

Added:  Topics covered SHOULD include:

     .  Requirements for recognizing trusted roles

     .  Responsibilities for each role (Examples of trusted roles
        include system administrators, security officers, and system
        auditors.)

     .  For each task identified, the number of individuals required
        to perform the task (n out of m rule) for each role.

     .  Identification and authentication requirements for each role

     .  When role separation is employed and how it is effected (if
        employed)


5.4.1. Types of events recorded -- 2nd sentence

From:  Audit records SHOULD include the date, time, responsible user
or process, and summary content data relating to the event.
To:    Audit records MUST include the date, time, responsible user
or process, and summary content data relating to the event.
Also added the following auditable events:

  . Key generation

  . Software and/or configuration updates to the CA

  . Clock adjustments


5.8. CA or RA termination

From: In the RPKI, each subscriber acts as a CA authoritative for the
   specified INRs that were distributed to that entity. Procedures
   associated with the termination of a CA MUST be described in the CPS
   for that CA.

To:  In the RPKI, each subscriber acts as a CA authoritative for the
   specified INRs that were distributed to that entity. Procedures
   associated with the termination of a CA MUST be described in the CPS
   for that CA. These procedures MUST include a provision to notify
   each entity that issued a certificate to the organization that is
   operating the CA that is terminating.

   Since the RA function MUST be provided by the same entity operating
   as the CA (see Section 1.4.2), there are no separate stipulations
   for RAs.

6.1.1. Key pair generation

Added sentence in the middle of the paragraph:
   (The CA has to check the quality of the keys only if it generates
   them (see 6.1.6)).


6.1.6. Public key parameters generation and quality checking

From: The public key parameters used in the RPKI are specified in RFC ZZZZ
   [RFCzzzz]. Each subscriber is responsible for performing checks on
   the quality of its key pair. CAs are not responsible for performing
   such checks for subscribers.

To:   The public key parameters used in the RPKI are specified in RFC ZZZZ
   [RFCzzzz]. Each subscriber is responsible for performing checks on
   the quality of its key pair. A CA is not responsible for performing
   such checks for subscribers except in the case where the CA
   generates the key pair on behalf of the subscriber.


6.3.2. Certificate operational periods and key pair usage periods

From: The INRs held by RIRs and the NIRs periodically change because these
   registries receive new distributions. Therefore it is appropriate for
   their certificates to hae lifetimes that match the periodicity of these
   changes

To:   The INRs held by a CA may periodically change because it receives
   new distributions. Therefore it is appropriate for its certificate to
   hae a lifetime that matches the periodicity of these changes.


9.12.2. Notification mechanism and period

From: The IESG MUST provide at least a six month advance notice of a
   change to this CP.

To:   Successive versions of the CP will be published with the statement
   "This CP takes effect on MM/DD/YYYY." MM/DD/YYYY MUST be a minimum
   of 6 months from the date of publication.


10. Security Considerations

Added: Since name uniqueness within the RPKI cannot be guaranteed, there is
   a risk that two or more CAs in the RPKI will issue certificates and
   CRLs under the same Issuer name. Path validation implementations
   that conform to the resource certification path validation algorithm
   [see RFCyyyy] verify that the same key was used to sign both the
   target (the resource certificate) and the corresponding CRL. So a
   name collision will not change the result. Use of the basic X.509
   path validation algorithm, which assumes name uniqueness, could
   result in a revoked certificate being accepted as valid or a valid
   certificate being rejected as revoked. Relying parties must ensure
   that the software they use to validate certificates issued under
   this policy verifies that the same key was used to sign both the
   certificate and the corresponding CRL, as specified in [RFCyyyy].


13.1. Normative References

Added:

[RFC2026]   Bradner, S., "The Internet Standards Process - Revision 3,"
       BCP 9, RFC 2026, October 1996.

[RFC3779]   Lynn, C., Kent, S., Seo, K., "X.509 Extensions for IP
       Addresses and AS Identifiers," RFC 3779, June 2004.


13.2. Informative References

Added:

[PROV] Huston, G., Loomans, R., Ellacott, B., Austein, R., "A Protocol
       for Provisioning Resource Certificates," work in progress.

[RFC3647]   Chokhani, S., Ford, W., Sabett, R., Merrill, C., Wu, S.,
       "Internet X.509 Public Key Infrastructure Certificate Policy and
       Certification Practices Framework," RFC 3647, November 2003.


_______________________________________________
sidr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sidr

Reply via email to