Agree with Patrice. The initial value was coming from the RFC7432 spec:
“An Ethernet Tag ID is a 32-bit field containing either a 12-bit or 24-bit identifier that identifies a particular broadcast domain (e.g., a VLAN) in an EVPN instance.” And now there are a few implementations out there that may check that the eth-tag is 24 bits and not larger. I prefer low order 24-bits rather than right aligned… Thanks. Jorge On 2/22/17, 3:39 AM, "BESS on behalf of Patrice Brissette (pbrisset)" <[email protected] on behalf of [email protected]> wrote: Himanshu, I don’t think we should make the Eth-tag a MUST be 24 bit. It should be MAY but if you decide to use 24 bits, it MUST be right aligned. Regards, Patrice Brissette On 2017-02-21, 2:51 PM, "Shah, Himanshu" <[email protected]> wrote: ‘MAY’ does not work. It has to be ‘MUST’, IMO. Thanks, Himanshu On 2/21/17, 2:22 PM, "BESS on behalf of Patrice Brissette (pbrisset)" <[email protected] on behalf of [email protected]> wrote: Folks, Why don’t we simply mention that the Eth–Tag is a 32 bit value and MAY be set to a 24 bits instance When 24 bits value is used is MAY be right aligned. Regards, Patrice Brissette On 2017-02-21, 2:18 PM, "BESS on behalf of Sami Boutros" <[email protected] on behalf of [email protected]> wrote: Hi John, I can add that the value is from 0 to 0x00ffffff, will that work? Thanks, Sami On 2/21/17, 10:56 AM, "John E Drake" <[email protected]> wrote: >Sami, > >Snipped, comment inline > >Yours Irrespectively, > >John > >> > >> >> Ethernet Tag ID 32-bit field MUST be set to the 24-bit VPWS service instance >> identifier value." >> > >> > >> > >> >Ok, but you still didn’t mention how the 24-bit value is to be aligned in the 32- >> bit field. I’m guessing there will be some 0-padding, but will that the at the >> beginning or the end? >> > >> >> I made the VPWS service instance identifier a 32-bit value in the new draft. >> > >[JD] I don't think you can do this as there are multiple implementations that use 24 bits > _______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess _______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess _______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess _______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess
