Hi Rob,
I believe rev-12 addresses the 3 DISCUSS points below.
We still have to do further updates to the document.
Regards,Reshad.
    On Monday, December 12, 2022, 12:03:19 PM EST, Robert Wilton via 
Datatracker <[email protected]> wrote:  
 
 Robert Wilton has entered the following ballot position for
draft-ietf-bfd-unsolicited-11: Discuss

When responding, please keep the subject line intact and reply to all
email addresses included in the To and CC lines. (Feel free to cut this
introductory paragraph, however.)


Please refer to 
https://www.ietf.org/about/groups/iesg/statements/handling-ballot-positions/ 
for more information about how to handle DISCUSS and COMMENT positions.


The document, along with other ballot positions, can be found here:
https://datatracker.ietf.org/doc/draft-ietf-bfd-unsolicited/



----------------------------------------------------------------------
DISCUSS:
----------------------------------------------------------------------

Hi,

Thanks for this document.

Please see my comments below for more details, but I'm balloting discuss on 3
points: (1) The document is somewhat unclear as to whether the configuration is
applied hierarchically (I presume that it is, if not then my second discuss
point is not valid and can be ignored). (2) As specified, I don't think that
the hierarchical configuration will work, because the interface level leaf
"defaults" will override an explicit value configured globally.  I.e.,
logically, the interface level leaf, if in scope, will always have a value. (3)
The document should provide an instance-data example in the appendix to
illustrate the use of this configuration.

Regards,
Rob


----------------------------------------------------------------------
COMMENT:
----------------------------------------------------------------------

Moderate level comments:

(1) p 3, sec 2.  Procedures for Unsolicited BFD

  When the passive side receives a BFD Control packet from the active
  side with 0 as "Your Discriminator" and does not find an existing BFD
  session, the passive side MAY create a matching BFD session toward
  the active side, if permitted by local configuration and policy.

I'm surprised that this is only a MAY and not a SHOULD or MUST.  I.e., if the
local configuration & policy allows passive BFD sessions why would they not be
created?

(2) p 4, sec 4.1.  Unsolicited BFD Hierarchy

  *  Globally, i.e. for all interfaces.  This requires support for the
      "unsolicited-params-global" feature.
  *  For specific interfaces.  This requires support for the
      "unsolicited-params-per-interface" feature.

>From this description, it is unclear to me whether the features enabling global
or per-interface configuration are meant to be an either/or (in which case, the
constraint could probably be expressed in the features), or whether a server is
allowed to support configuration both globally and override the global
configuration with interface specific configuration.  My subsequent discuss
comments assume the latter.  Either way, it would be helpful for this text in
this section (and probably the YANG module) to be a bit more explicit on this
regard.

(3) p 8, sec 4.2.  Unsolicited BFD Module

      augment "/rt:routing/rt:control-plane-protocols/"
            + "rt:control-plane-protocol/bfd:bfd/bfd-ip-sh:ip-sh/"
            + "bfd-ip-sh:interfaces" {
        if-feature bfd-unsol:unsolicited-params-per-interface;
        description
          "Augmentation for BFD unsolicited on IP single-hop interface";
        container unsolicited {
          description
            "BFD IP single-hop interface unsolicited top level
            container";
          leaf enabled {
            type boolean;
            default false;

I'm not sure that you want a default value specified in the YANG here since
this would have precedence over any explicitly configured global default value.

(4) p 8, sec 4.2.  Unsolicited BFD Module

            description
              "BFD unsolicited enabled on this interface.";
          }
          uses bfd-types:base-cfg-parms;

You have the same issue here as above, in that the default values directly
associated with the leaves in this grouping will always take precedence over
any configured global value.  I.e., the configuration, if properly implemented
cannot be hierarchical.  The pragmatic solution is to copy the used grouping
inline here and delete the default statements.  This has the advantage that the
descriptions can also make the hierarchical behavior of the configuration
explicit.

(5) p 9, sec 4.2.  Unsolicited BFD Module

    augment "/rt:routing/rt:control-plane-protocols/"
          + "rt:control-plane-protocol/bfd:bfd/bfd-ip-sh:ip-sh/"
          + "bfd-ip-sh:sessions/bfd-ip-sh:session" {
      description
        "Augmentation for BFD unsolicited on IP single-hop session";
        leaf role {
          type bfd-unsol:role;
          config false;
          description
            "Role of local system during BFD session initialization.";
        }
    }
  }
  <CODE ENDS>

Please add an instance data example to an appendix to illustrate the use of
this YANG model.  This helps readers and can further emphasize the hierarchical
nature of the configuration.

Minor level comments:

(6) p 3, sec 2.  Procedures for Unsolicited BFD

  Passive unsolicited BFD support MUST be disabled by default, and MUST
  require explicit configuration to be enabled.  On the passive side,
  the desired BFD parameters SHOULD be configurable.  The passive side
  MAY also choose to use the parameters that the active side uses in
  its BFD Control packets.  The "My Discriminator", however, MUST be
  chosen to allow multiple unsolicited BFD sessions.

Rather then configured values on the passive side, did the authors consider
setting minimum configuration limits?  E.g., rather than define desired
send/receive limits, instead, configure lower bounds on what the minimum tx
interval may be (to prevent DDOS attacks).

Nit level comments:

(7) p 3, sec 2.  Procedures for Unsolicited BFD

  The passive side MUST then start sending BFD Control packets and
  perform the necessary procedure for bringing up, maintaining and
  tearing down the BFD session.  If the BFD session fails to get
  established within certain specified time, or if an established BFD
  session goes down, the passive side SHOULD stop sending BFD Control
  packets and MAY delete the BFD session created until BFD Control
  packets are initiated by the active side again.

Nit, within certain specified => within a specified

(8) p 6, sec 4.2.  Unsolicited BFD Module

        Copyright (c) 2021 IETF Trust and the persons
        identified as authors of the code.  All rights reserved.

This copyright statement will need to be fixed.



  

Reply via email to