Stewart, Respect your advice. However, some wording in the proposed charter are too ambiguous, is it the intent? For example: "An NVO3 solution (known here as a Data Center Virtual Private Network (DCVPN)) is a VPN that is viable across a scaling range of a few thousand VMs to several million VMs running on greater than 100K physical servers." Do you mean "one VPN across a scaling range of million VMs"? or many VPNs combined to scale range of million VMs? Another example: "NVO3 will consider approaches to multi-tenancy that reside at the network layer rather than using traditional isolation mechanisms that rely on the underlying layer 2 technology (e.g., VLANs)" "network layer" can mean different things to different people. Why not simply say "NV03 will consider approaches to multi-tenancy which do not rely on Layer 2 VLANs"? 3rd example: " The NVO3 WG will determine which types of service are needed by typical DC deployments" Data center provide Computing and storage services. Network facilitates the connection among Computing entities and storage entities. Why does NV03 WG need to determine what types of "services" are needed by typical DC deployment? Do you mean "NV03 WG will consider network deployment by typical DC"? Linda Dunbar > -----Original Message----- > From: Stewart Bryant [mailto:stbryant@xxxxxxxxx] > Sent: Tuesday, April 24, 2012 4:53 PM > To: Linda Dunbar > Cc: nvo3@xxxxxxxx; iesg@xxxxxxxx; IETF Discussion > Subject: Re: Key difference between DCVPN and L2VPN/L3VPN > > On 24/04/2012 22:39, Linda Dunbar wrote: > > I think that the Charter should have some text to identify the key > differences between DCVPN and L2VPN/L3VPN. If the key differences > aren't properly described, all the protocols developed for L2VPN and > L3VPN will pop up as potential solutions. > I do not think that the charter is the place to provide a technical > definition of the difference between DCVPN and L2VPN/L3VPN. The charter > sets the scope of the work. > > > > > Here are some key differences I see: > > - Majority of VMs (or End-Systems) do communicate with external peers, > even though their volume may not be large. Most external peers access > their VMs (or End-Systems) in Data Centers via internet or IP network > until packets reach the Data Center gateway. > > > > - Majority of VMs are end systems, vs. nodes in L2VPN/L3VPN are not. > > > That really would be a better fit in the requirements and framework > documents. > > - Stewart >