As highlighted before a simplified architecture results in a smarter data center solution. Taking an architectural approach that allows the enterprise IT to use common tools, products, protocols and standards across the whole network infrastructure reduces both cost and risk while providing a much greater level of flexibility and agility to deliver applications from the virtualized data center down to the mobile user.
So if we look at our architecture we find the following solution layers – from server to the fabric:•
- Virtualization, Service Insertion
- Visibility and automation when virtual servers and new services are provisioned, moved, retired
- Data Center Bridging
- Supporting DC I/O and SAN convergence
- Virtual Switching
- Increasing BW and enabling resilient links to servers/blade center switches
- Fabric Core Mesh
- Scale DC core & lower latency L2&L3
- Data Center Interconnect
- Optimize Routing into multiple DC and create a single (L2) domain
- Application Awareness
- Enabling application visibility and control for DC optimization
In the virtualization solution layer we focus on the orchestrated provisioning of virtual and physical fabric. We leverage our policies and policy management framework for this that we are using at the wired and wireless campus edge for years. So no need to develop a new solution for VM workloads in the data center. And voila: you end up with same management and control solution – OneFabric Control Center with Data Center Manager DCM – from Data Center to Edge.
Data center bridging applies to the data center – but the products that we support this on are also the ones we use in the campus infrastructure. So no need for a dedicated set of products – makes life easier again.
Virtual switching via our VSB virtual switch bonding solution allows us to provide both a resilient link to servers and blade centers in the data center but it is also beneficial to increase the bandwidth via link aggregation – and so active/active links – into the wiring closets without upgrading the link speed from 1G to 10G for example. Simple again and the same across the product line. No need to train you staff for different technologies.
The fabric core mesh scales beyond a VSB implementation. Next generation core need to support an active/active configuration while providing a high degree of resiliency that:
- Contains failures so only directly affected traffic is impacted during restoration
- Enables rapid restoration of broadcast and multicast connectivity
- Leverages all of the available physical connectivity, with no lost bandwidth
- Enables fast restoration of connectivity after failure
There are two competing standards in the process of ratification that will increase the resiliency of tomorrow’s data center fabrics:
- Shortest Path Bridging (SPB) – IEEE 802.1aq work group
- Transparent Interconnect of Lots of Links (TRILL) – IETF TRILL work group
Each of these standards is aimed at simplifying the network topology and providing an active mesh between the edge and core of data center networks.
IETF TRILL and IEEE SPB both use IS-IS as the routing protocol to achieve similar goals. The IEEE has committed itself to supporting all existing and new IEEE standards (particularly the IEEE Data Center Bridging protocols, but also existing management protocols, Ethernet IEEE 802.1ag (OAM), etc.) via IEEE SPB.
The Enterasys Data Center Fabric 2.0 will use IEEE SPB because of the leverage of existing standards and the fact that it builds upon and is fully interoperable with the existing data center as well as campus LANs running MSTP and it will improve the resiliency of tomorrow’s networks. Due to its nature as an IEEE standard it allows for the migration from existing infrastructures with little to no disruptions and can be used end to end in a enterprise network. Again no need to re-invent the wheel here.
Additionally Fabric Routing as part of a core mesh solution is a mechanism to provide distributed routing integrated in SPB and RSTP/MSTP/LACP based switch/routers to address the need for maximum throughput, lowest latency and optimized traffic flows inside the data center fabric. In the context of both the LAN and the WAN, north-south traffic is the client server traffic that goes between users in a branch office and the data center that hosts the application that they are accessing. In the context of the data center, east-west traffic (which makes up almost 80% of the traffic in the DC) is the traffic that goes between servers in a given data center fabric. Fabric Routing is a unique innovation by Enterasys built upon and interoperable with VRRP so administrators can leverage their existing knowledge for the implementation. It can be also applied and provides value in the campus LAN to create active/active VRRP configurations.
I will follow up with two dedicated blogs on DCI – but the same scheme applies here.