Hi Martin,

We have just implemented RFC7814 a couple of months before in response to the 
demands of some of our customers, and those customers have deployed this 
L3-based overlay technology within their hyper-scale cloud data center network 
environment recently. We are planning to implement 
draft-ietf-bess-virtual-subnet-fib-reduction in the next step.

I have just noticed the following statement from 

"... All major clouds, including Amazon EC2, Amazon VPC, Google Compute Engine 
and Microsoft Azure, allow only unicast datagrams with IP payloads. Broadcast 
datagrams and non-IP payloads are not allowed (with very limited exceptions to 
make parts of the essential ARP and DHCP protocols work). "

IMHO, when the L3 overlay technology is more widely deployed in hyper-scale 
data centers, the on-demand FIB installation mechanism as described in 
draft-ietf-bess-virtual-subnet-fib-reduction would become significantly 
important to the operators of those hyper-scale data centers . Once this 
mechanism becomes an IETF standard, those operators would ask their vendors to 
implement it. In fact, this is the case of our implementation of RFC7814. 
Hence, I strongly suggest to move this to IESG.

BTW, since it has become a fact that various data plane encapsulation schemes 
could be used for the L3VPN solution, hence I wonder whether the implementation 
of L3 Conversational Learning of DFA as described on page 36 of 
(http://www.valleytalk.org/wp-content/uploads/2013/08/ciscoDFA.pdf) could be 
looked as an existing implementation:)

Best regards,

> -----邮件原件-----
> 发件人: BESS [mailto:bess-boun...@ietf.org] 代表 Martin Vigoureux
> 发送时间: 2016年10月11日 23:01
> 收件人: BESS
> 主题: [bess] Lack of implementation for
> draft-ietf-bess-virtual-subnet-fib-reduction - submit to IESG?
> WG,
> we have recently LCed draft-ietf-bess-virtual-subnet-fib-reduction and there 
> was
> sufficient support to move forward. However we haven't received any input on
> existing implementations.
> As per [1] we are thus asking the WG whether we should nevertheless move this
> to IESG or wait until implementations exist.
> Please respond. Thank you.
> M&T
> [1]:
> https://mailarchive.ietf.org/arch/msg/bess/cG3X1tTqb_vPC4rg56SEdkjqDpw
> _______________________________________________
> BESS mailing list
> BESS@ietf.org
> https://www.ietf.org/mailman/listinfo/bess
BESS mailing list

Reply via email to