What Exactly Is The Most Significant Technological Innovation Point In Ccie Sp?- Essential Details That Most Engineers Must Take Into Consideration

If we are working on ccie sp lab, MPLS is an essential point we will have to study.

A lot of engineers are worried with precisely what seems to be frivolous generation of innovative encapsulation formats supporting virtual networks. Whilst STT helps make technical sense, it's harder to determine the benefits associated with VXLAN together with NVGRE. Scott Lowe wrote a great blog post recently exactly where he asked an extremely valid inquiry: "Couldn't we use MPLS over GRE or IP?" We could, but we wouldn't gain anything by doing that.

CCIE SP WORKBOOK: RFC 4023 defines two techniques of MPLS in IP encapsulation: MPLS label stack over IP (using IP protocol 137) and then MPLS label stack over GRE (using MPLS protocol type in GRE header). We can use either one of these and use either the typical MPLS semantics or improper use MPLS label as virtual network identifier (VNI). Let's study both possibilities.

Misusing MPLS label as VNI

In theory, one might use MPLS over IP or maybe MPLS over GRE instead of VXLAN (or NVGRE) and make use of the first MPLS label as the VNI. Despite the fact that this could work (of course, NVGRE reuses GRE key as VNI), it would not gain us anything. The current tools might not recognize this "creative" use of MPLS labels, and we even wouldn't have the control plane and will have to trust in IP multicast to emulate virtual network L2 flooding.

CCIE SERVICE PROVIDER WORKBOOK : The MPLS label = VNI technique could be totally incompatible with existing MPLS stacks and will thus require new software in virtual-to-physical gateways. It can also go against the gist of MPLS - labels needs to have local significance (while VNI has network-wide importance) and must be assigned individually by individual MPLS nodes (egress PE-routers in MPLS/VPN scenario).

It's also questionable whether or not the existing hardware could process MAC in MPLS in GRE in IP packets, which might be the only probable benefit of this strategy. I know that some (expensive) linecards in Catalyst 6500 could process IP in MPLS in GRE packets (as conduct some switches from Juniper and HP), but should it process MAC in MPLS in GRE? Who knows.

Finally, just like NVGRE, MPLS-over-GRE or MPLS over IP framing with MPLS label being used as the VNI lacks entropy which may be used for load balancing functions; existing switches might not be able to load balance traffic between two hypervisor web-hosts unless every hypervisor hosts will use multiple IP addresses.

Employing MPLS over IP/GRE to implement virtual networks helps make marginal sense, will not solve the load balancing problems NVGRE is dealing with, and needs significant investment in the hypervisor-side control plane if you want to do it right. I don't expect to find it applied any time soon.