Re: [pim] draft-ietf-pim-igmp-mld-snooping-yang-01 comments

[Stig Venaas <[e-mail]>]

Hi

Thanks, I do think it should be 1..7 since 1 is a valid value. It
would have been different if the RFC said MUST NOT.

Thanks for making this and other changes. I hope the same change can
be made in the IGMP/MLD (not snooping) model.

Stig


On Mon, Apr 23, 2018 at 7:59 PM, Hongji Zhao <[e-mail]> wrote:
> Hi Stig,
>
> Thanks a lot for your comments.
>
>
>
> Regarding your main comment about the robustness variable. I have checked
> the RFC 3376
>
>
>
> Quote:
>
> 8.1. Robustness Variable
>
> The Robustness Variable allows tuning for the expected packet loss on
>
> a network. If a network is expected to be lossy, the Robustness
>
> Variable may be increased. IGMP is robust to (Robustness Variable -
>
> 1) packet losses. The Robustness Variable MUST NOT be zero, and
>
> SHOULD NOT be one. Default: 2   ------------------------Here it says the
> value should not be one. So I set the range for the robustness variable
> between 2 and 7.  I will modify the range to 1..7. Is it ok?
>
>
>
>
>
> I will check other comments and revise the draft later on.
>
>
>
>
>
> BR/Hongji
>
> 赵宏吉
>
>
>
> -----Original Message-----
> From: Stig Venaas [mailto:[e-mail]]
> Sent: Tuesday, April 24, 2018 6:06 AM
> To: [e-mail]; [e-mail]
> Subject: draft-ietf-pim-igmp-mld-snooping-yang-01 comments
>
>
>
> Hi
>
>
>
> After reading through the draft I have some comments.
>
>
>
> The main comment I have is regarding the robustness variable. I see it says
> the value must be in the range 2..7. But RFC 3376 does allow 1,
>
> quote:
>
>
>
>    The Robustness Variable allows tuning for the expected packet loss on
>
>    a network.  If a network is expected to be lossy, the Robustness
>
>    Variable may be increased.  IGMP is robust to (Robustness Variable -
>
>    1) packet losses.  The Robustness Variable MUST NOT be zero, and
>
>    SHOULD NOT be one.  Default: 2
>
>
>
> Shouldn't this document, and also the IGMP-MLD model allow the value 1?
>
>
>
> Below are various less important comments.
>
>
>
>
>
> In section 2:
>
>    The goal of this document is to define a data model that provides a
>
>    common user interface to IGMP and MLD Snooping.  There is very
>
>    information that is designated as "mandatory", providing freedom for
>
>    vendors to adapt this data model to their respective product
>
>    implementations.
>
>
>
> It seems something is wrong with this sentence "There is very
> information...". Are there some words missing?
>
>
>
> In 2.1, I'm not sure what this sentence means:
>
>    The YANG module includes IGMP and MLD Snooping instances
>
>    definition, instance references in the scenario of BRIDGE, VPLS.
>
> Note that in 2.3 it says "bridge, L2VPN".
>
>
>
> In 2.1 it says "clearing the specified IGMP and MLD Snooping". I think it
> should say something more, perhaps "IGMP and MLD Snooping group tables".
>
>
>
> In 2.1, 2 dots at the end "data modeling language..".
>
>
>
> In 2.2, double space "to configure IGMP and  MLD".
>
>
>
> In 2.2, maybe find a better way to phrase "The attribute who could be read
> and written shows configuration data."
>
>
>
> Section 3 comments.
>
>
>
> It says "fowarding" several places.
>
>
>
> It says "Outgoing ac in l2vpn fowarding". Should "ac" be "AC"? Same question
> regarding "pw". Also, should "mac"
>
> be "MAC" in the descriptions? Also "l2vpn" should be "L2VPN" I think.
>
>
>
> It says "VSI" in one place. What is VSI?
>
>
>
> "leaf require-router-alert", is it to set RA, and/or requiring RA on
> received messages.
>
>
>
> I think "The time after the device created L2 multicast record."
>
> should be "The time elapsed since".
>
>
>
> For last reporter it says "The last host address which has sent the
> report"... It should say "address of the last host which has.."
>
> or perhaps better "address of the last report received...".
>
>
>
> What is the name/key used for a snooping instance? Can a user choose an
> arbitrary name, or do implementations somehow decide the name?
>
>
>
> In many descriptions it says "IP address", but some places it says "IPv4
> address" or "IPv6 address". I think it is best to include the version in the
> description when the leaf is for a particular version.
>
>
>
> Section 4 Security considerations probably need to be expanded. At the least
> explain why no issues if that's the case.
>
>
>
> Section 5 needs to be updated, there is some work for IANA.
>
>
>
> Mahesh's affiliation and address need to be updated.
>
>
>
> Stig
>
>
>
>
>
> BR/Hongji
>
> 赵宏吉
>
>