Re: [OPSEC] RtgDir: Last Call Review of draft-ietf-opsec-v6-21.txt - "Operational Security Considerations for IPv6 Networks"

["Eric Vyncke (evyncke)" <[e-mail]>]

Acee, 

I appreciate the quick reply of yours ;-)

About, a shared ACL between IPv4 and IPv6, if the YANG model has one separate ACL per address family, then it is broken (<AD hat> I was not yet the INT AD at that time </AD hat>). 

Most of the firewalls (also a layer-3 device) have a single ACL for both address families and now, let's not conflate the actual TCAM implementation (that is obviously per address family) with the CLI or management interfaces that SHOULD (really) be protocol agnostic.

Anyway, you kindly wrote " However, I don't feel that strongly." ;-)

Regards,

-éric

-----Original Message-----
From: "Acee Lindem (acee)" <[e-mail]>
Date: Tuesday, 9 February 2021 at 12:28
To: Eric Vyncke <[e-mail]>, "[e-mail]" <[e-mail]>
Cc: Routing Directorate <[e-mail]>, "[e-mail]" <[e-mail]>, "[e-mail]" <[e-mail]>
Subject: Re: [OPSEC] RtgDir: Last Call Review of draft-ietf-opsec-v6-21.txt - "Operational Security Considerations for IPv6 Networks"

    Hi Eric,

    Thanks for incorporating my comments. See one inline. 

    On 2/9/21, 6:10 AM, "Eric Vyncke (evyncke)" <[e-mail]> wrote:

        Hello Acee,

        Thank you for your directorate review and sorry for such a belated reply!

        Special thanks for the DIFF containing suggestions for improving the text. Most of them have been applied (none of the authors is English native so such assistance is welcome)

        Look below for EV> for more comments.

        Regards

        -éric

        -----Original Message-----
        From: "Acee Lindem (acee)" <[e-mail]>
        Date: Tuesday, 3 December 2019 at 17:21
        To: "[e-mail]" <[e-mail]>, "[e-mail]" <[e-mail]>
        Cc: Routing Directorate <[e-mail]>, "[e-mail]" <[e-mail]>
        Subject: [OPSEC] RtgDir: Last Call Review of draft-ietf-opsec-v6-21.txt - "Operational Security Considerations for IPv6 Networks"

             Hello,

            I have been selected as the Routing Directorate reviewer for this draft.
            The Routing Directorate seeks to review all routing or routing-related
            drafts as they pass through IETF last call and IESG review, and
            sometimes on special request. The purpose of the review is to provide
            assistance to the Routing ADs. For more information about the Routing
            Directorate, please see ​

              http://trac.tools.ietf.org/area/rtg/trac/wiki/RtgDir

            Although these comments are primarily for the use of the Routing ADs,
            it would be helpful if you could consider them along with any other
            IETF Early Review/Last Call  comments that you receive, and strive to
            resolve them through discussion or by updating the draft.

            Document: draft-ietf-opsec-v6-21.txt
            Reviewer: Acee Lindem
            Review Date: 12/2/2019
            IETF LC End Date: Soon
            Intended Status:  Informational

            Summary: The document contains a lot of useful recommendations and
                     references for Operational Security in IPv6 networks. Since
                            the document has "Informational" status, none of the text is
                            normative.

                            While the information content is very good, parts of the
                            document are very hard to read and need revision. In general,
                            the usage of long clauses connected by semicolons should be
                            discouraged and the lists connected in this manner should
                            be replaced with complete sentences. I've attached a diffs
                            with editorial suggests but didn't try and rewrite all the
                            semicolon connected text segments.

        EV> let's hope that the RFC Editor will find and remediate those long constructs.

                            There are also minor issues that need to be addressed.

            Major Issues: None

            Minor Issues:

                1. Section 1.0 - What do you mean by "updating it with that have been
                   standardized since 2007."? It just doesn't read right.

        EV> text has been simplified in -23

                2. Section 2.1 - IPv4 also allows multiple addresses per interface,
                   i.e., secondary addresses. So what is new?

        EV>  last sentence now reads as
         "Having by default multiple
           IPv6 addresses per interface is a major change compared to the unique
           IPv4 address per interface for hosts (secondary IPv4 addresses are
           not common); especially for audits (see section Section 2.6.2.3)."

                3. Section 2.1.5 - The whole discussion on how to use Router
                   Advertisement (RA) messages lacks enough context to understand.
                   Also, expand RA in the first occurrence.

        EV> text was not clear indeed, changed

                4. Section 2.2.3 - Expand out NDP since it is not clear that it is
                   Neighbor Discovery Protocol from the context. It is expanded later
                   in section 2.3.

        EV> thanks, fixed

                5. Section 2.4 - RFC 6192 not only defines the "router control plane"
                    but provides much better guidance for control plane filtering than
                    section 2.4.1 and 2.4.2.

        EV> text updated

                6. Section 2.4.1 and 2.4.2 - The ingress ACL should only be applied on
                   the packets punted to the RP.

        EV> indeed, added

                7. Section 2.4.1 - If OSPFv3 vitual links are used, the destination
                   address will not be a link-local address.

        EV> trusting you on this one, text modified

                8. Section 2.4.3 - Suggest references for Path MTU Discovery
                   and traceroute.

        EV> good idea

                9. Section 2.5.1 - HMAC MD5 is considered vulnerable.

        EV> let's indeed remove this paragraph

               10. Section 2.5.2 - What prior section describes the operational
                   costs of IPsec?

        EV> oups the previous section was deleted revisions ago...

               11. Section 2.5.3 - Need expansion and reference for RADB.

        EV> indeed, added reference to https://www.radb.net/ 

               12. Section 2.6 - Need expansion and reference for GDPR.

        EV> indeed, added reference to https://eur-lex.europa.eu/eli/reg/2016/679/oj

               13. Section 2.7.1 - ACLs are typically per address family so this
                   recommendation isn't
                   really feasible. Please revise.

        EV> I disagree, this is a platform limitation. Text unchanged.

    ACEE> I don't know of a platform that doesn't have this limitation. Also, the IETF YANG model is organized with all ACL entries in an ACL being the same type. 

    See excerpt from RFC 8519:

      list acl {
          key "name";
          description
            "An ACL is an ordered list of ACEs.  Each ACE has a
             list of match criteria and a list of actions.
             Since there are several kinds of ACLs implemented
             with different attributes for different vendors,
             this model accommodates customizing ACLs for
             each kind and for each vendor.";
          leaf name {
            type string {
              length "1..64";
            }
            description
              "The name of the access list.  A device MAY further
               restrict the length of this name; space and special
               characters are not allowed.";
          }
          leaf type {
            type acl-type;
            description
              "Type of ACL.  Indicates the primary intended
               type of match criteria (e.g., Ethernet, IPv4, IPv6, mixed,
               etc.) used in the list instance.";
          }

    However, I don't feel that strongly. 

    Thanks,
    Acee


               14. Section 2.7.2.6 - Expand MAP-E and MAP-T.

        EV> done

               15. Section 3.1 and 4.1 - Define bogon and provide reference.

        EV> done, added reference to CYMRU

               16. Section 3.2 - Bad reference in fourth paragraph.

        EV> fixed in -22

               17. Section 5 - Suggest references for Teredo tunnels and NAT-PT.
                   Also, expand NAT-PT on first occurrence.

        EV> good idea for Teredo (added) and NAPT is now defined in section 1

            Nits: Attached diff with suggested edits.

        EV> big thanks for them

            Thanks,
            Acee