This is my 3rd review of the document. While I have focused on changes done since my last review, some comments apply to parts of the YANG model which hasn't changed recently. Main comments ============= - Feature immediate-leave mentions “fast leave” in the description. RFC3376 mentions fast leave but not immediate leave. Should the feature, and the leaf node which depends on it, be renamed to fast-leave? - Leaf send-query, description says that it cooperates with parameter querier-source. I believe there should be enforcement that send-query can only be set if querier-source is also set (must statement)? Also querier-source in IGMP mentions VLAN, no such mention in MLD, is that correct? Questions ======== - Feature static-l2-multicast-group. The description mentions L2 multicast static-group. Is it a static multicast group or a multicast static group? I believe it’s the former and the description should be changed? - igmp-version and mld-version are both uint8 with a range (1..3 and 1..2). If we get a new version someday, the range will have to be changed. I don’t recall if this was brought up before. Another option is to use an identity. I realize we don’t spin out a new version frequently so this may not be an issue. And probably with a new version other changes would be needed anyway… - Last-reporter is present under the group and also under each source entry. Is the one under the group the last host from all sources? - Leaf node require-router-alert. What happens if it’s set to true and the IP hdr does not contain RA? Consider updating description and/or adding a reference. Minor ===== - Typo in bridge-router-interface: dynamicly. - No need to mention the leaf name in description, e.g. in l2-service-type - Leaf “host-address”, rename to “address” since the list is called host (no need to duplicate host). Regards, Reshad.